# OpenClaw

Self-hosted gateway that connects Discord, Google Chat, iMessage, Matrix, Microsoft Teams, Signal, Slack, Telegram, WhatsApp, Zalo, and more to AI coding agents. Run one Gateway process on your own machine and message your AI assistant from anywhere.

> Use this file as a lightweight map of the OpenClaw documentation. Fetch individual pages as Markdown with `.md` URLs or `Accept: text/markdown`.

## Agent Resources

- [Markdown page export](https://docs.openclaw.ai/start/getting-started.md): Append `.md` to any docs page URL for clean Markdown.
- [Sitemap](https://docs.openclaw.ai/sitemap.xml): Search crawler URL index.
- [Robots policy](https://docs.openclaw.ai/robots.txt): Bot and crawler policy.

## Documentation Index

- [Agent runtime architecture](https://docs.openclaw.ai/agent-runtime-architecture): How OpenClaw structures the built-in agent runtime: code layout, boundaries, resource manifests, and runtime selection.
- [BlueBubbles removal and the imsg iMessage path](https://docs.openclaw.ai/announcements/bluebubbles-imessage): BlueBubbles support was removed from OpenClaw. Use the official iMessage plugin with imsg for new and migrated iMessage setups.
- [Auth credential semantics](https://docs.openclaw.ai/auth-credential-semantics): Canonical credential eligibility and resolution semantics for auth profiles
- [Automation](https://docs.openclaw.ai/automation): Overview of automation mechanisms: automations, hooks, standing orders, and workflows
- [Automations](https://docs.openclaw.ai/automation/cron-jobs): Automations: scheduled jobs, webhooks, and Gmail PubSub triggers for the Gateway scheduler
- [Automation delivery](https://docs.openclaw.ai/automation/cron-jobs/delivery): Announce, webhook, and none delivery modes, failure alerts, and output language
- [Gmail PubSub triggers](https://docs.openclaw.ai/automation/cron-jobs/gmail): Wire Gmail inbox events into OpenClaw through Google Pub/Sub and a restricted reader
- [How automations work](https://docs.openclaw.ai/automation/cron-jobs/how-it-works): Automation runtime model, run lifecycle, and promoting a repeated job into a schedule
- [Manage automations](https://docs.openclaw.ai/automation/cron-jobs/managing-jobs): CLI examples, job management commands, run history, and cron configuration keys
- [Automation payloads](https://docs.openclaw.ai/automation/cron-jobs/payloads): Payload kinds, agent-turn flags, command and script payloads, and session execution styles
- [Automation schedules](https://docs.openclaw.ai/automation/cron-jobs/schedules): Schedule kinds, cron expressions, stream sources, pacing, and condition triggers
- [Automation troubleshooting](https://docs.openclaw.ai/automation/cron-jobs/troubleshooting): Command ladder and fixes for jobs that do not fire or do not deliver
- [Inbound webhooks](https://docs.openclaw.ai/automation/cron-jobs/webhooks): Gateway HTTP hooks that let an external service wake an agent or submit a turn
- [Hooks](https://docs.openclaw.ai/automation/hooks): Internal hooks: install, write, and verify automation for commands and lifecycle events
- [Bundled hooks](https://docs.openclaw.ai/automation/hooks/bundled-hooks): The hooks shipped with OpenClaw and the behavior and options of each
- [Hook configuration and discovery](https://docs.openclaw.ai/automation/hooks/configuration): Selecting and enabling internal hooks, and how the Gateway discovers them across sources
- [Hook event types and context](https://docs.openclaw.ai/automation/hooks/event-types): Every internal event key, its trigger and wait behavior, and the context each producer supplies
- [Hook troubleshooting](https://docs.openclaw.ai/automation/hooks/troubleshooting): Diagnose a hook that is not discovered, not eligible, or not executing
- [Writing hooks](https://docs.openclaw.ai/automation/hooks/writing-hooks): Hook directory layout, the handler contract, reply delivery, and the HOOK.md metadata fields
- [IMAP email trigger](https://docs.openclaw.ai/automation/imap): Watch an IMAP mailbox and route authenticated incoming email to an isolated restricted reader agent
- [Standing orders](https://docs.openclaw.ai/automation/standing-orders): Define permanent operating authority for autonomous agent programs
- [Chat channels](https://docs.openclaw.ai/channels): Messaging platforms OpenClaw can connect to
- [A2A](https://docs.openclaw.ai/channels/a2a): Connect external agents to OpenClaw through the A2A 1.0 JSON-RPC protocol
- [Access groups](https://docs.openclaw.ai/channels/access-groups): Reusable sender allowlists for message channels
- [Ambient room events](https://docs.openclaw.ai/channels/ambient-room-events): Let supported group rooms provide quiet context unless the agent sends with the message tool
- [Bot loop protection](https://docs.openclaw.ai/channels/bot-loop-protection): Bot-to-bot loop protection defaults and channel overrides
- [Broadcast groups](https://docs.openclaw.ai/channels/broadcast-groups): Run bounded agent group threads across messaging channels
- [Buzz](https://docs.openclaw.ai/channels/buzz): Connect OpenClaw agents to Buzz rooms
- [Channel routing](https://docs.openclaw.ai/channels/channel-routing): How OpenClaw picks an agent and session for each inbound message, and how replies route back
- [ClickClack](https://docs.openclaw.ai/channels/clickclack): ClickClack bot-token channel setup and target syntax
- [Discord](https://docs.openclaw.ai/channels/discord): Discord bot setup, config keys, components, voice, and troubleshooting
- [Discord Activities](https://docs.openclaw.ai/channels/discord-activities): Launch self-contained OpenClaw HTML widgets inside Discord Activities
- [Discord access control](https://docs.openclaw.ai/channels/discord/access-control): DM policy, guild allowlists, mention gating, role routing, command auth, and action gates
- [Discord events and operations](https://docs.openclaw.ai/channels/discord/events): Reaction notifications, presence events and configuration, config writes, proxying, and PluralKit
- [Discord message behavior](https://docs.openclaw.ai/channels/discord/messaging): Runtime model, reply tags, link previews, live stream previews, ack reactions, and mention aliases
- [Discord components and approvals](https://docs.openclaw.ai/channels/discord/rich-messages): Components v2 containers, interactive component payloads, and exec approvals in Discord
- [Discord setup](https://docs.openclaw.ai/channels/discord/setup): Create the Discord application, invite the bot, and set up a guild workspace
- [Discord threads and sessions](https://docs.openclaw.ai/channels/discord/threads-and-sessions): Forum threads, history and context limits, subagent thread sessions, and ACP channel bindings
- [Discord troubleshooting](https://docs.openclaw.ai/channels/discord/troubleshooting): Fixes for blocked guild messages, intent errors, gateway timeouts, bot loops, and voice STT drops
- [Discord voice channels](https://docs.openclaw.ai/channels/discord/voice-channels): Realtime Discord voice channels, auto-join, realtime modes, and voice message attachments
- [Follow users in Discord voice](https://docs.openclaw.ai/channels/discord/voice-follow): Make the bot join, move between, and leave Discord voice channels with selected users
- [Discord voice transcripts and meeting notes](https://docs.openclaw.ai/channels/discord/voice-transcripts): Continuous voice capture, transcript storage, and automatic meeting notes
- [Feishu](https://docs.openclaw.ai/channels/feishu): Feishu bot overview, features, and configuration
- [Feishu access control](https://docs.openclaw.ai/channels/feishu/access-control): Feishu DM policy, group policy, mention gating, and how to find chat and user IDs
- [Feishu advanced configuration](https://docs.openclaw.ai/channels/feishu/advanced-configuration): Feishu multi-account setup, message limits, streaming, workspace tools, ACP sessions, and multi-agent routing
- [Feishu configuration reference](https://docs.openclaw.ai/channels/feishu/configuration-reference): Every `channels.feishu` configuration key with its default
- [Feishu dynamic agents](https://docs.openclaw.ai/channels/feishu/dynamic-agents): Per-user agent isolation for Feishu, giving each DM user their own workspace and agent
- [Feishu message types](https://docs.openclaw.ai/channels/feishu/messaging): Which Feishu message types OpenClaw receives and sends, including stickers and thread replies
- [Feishu setup](https://docs.openclaw.ai/channels/feishu/setup): Feishu/Lark bot setup wizard, and how inbound events are durably queued
- [Feishu troubleshooting](https://docs.openclaw.ai/channels/feishu/troubleshooting): Silent Feishu bots, missing events, QR setup failures, and leaked App Secrets
- [Google Chat](https://docs.openclaw.ai/channels/googlechat): Google Chat app support status, capabilities, and configuration
- [WhatsApp group messages](https://docs.openclaw.ai/channels/group-messages): WhatsApp group message handling — activation, allowlists, sessions, and context injection
- [Groups](https://docs.openclaw.ai/channels/groups): Group chat behavior across surfaces (Discord/iMessage/Matrix/Microsoft Teams/QQBot/Signal/Slack/Telegram/WhatsApp/Zalo)
- [iMessage](https://docs.openclaw.ai/channels/imessage): Native iMessage support via imsg (JSON-RPC over stdio), with private API actions for replies, tapbacks, effects, polls, attachments, and group management. Preferred for new OpenClaw iMessage setups when host requirements fit.
- [Coming from BlueBubbles](https://docs.openclaw.ai/channels/imessage-from-bluebubbles): Translate old BlueBubbles configs to the official iMessage plugin: key mapping, group allowlist gates, and cutover verification.
- [iMessage access control and routing](https://docs.openclaw.ai/channels/imessage/access-control): iMessage DM and group policy, mention gating, per-group prompts, ACP bindings, and config writes
- [iMessage deployment patterns](https://docs.openclaw.ai/channels/imessage/deployment): Dedicated bot macOS user, remote Mac over Tailscale, multi-account, and DM history patterns
- [iMessage media and attachments](https://docs.openclaw.ai/channels/imessage/media): iMessage attachment ingestion, outbound staging, text chunking, and delivery target formats
- [iMessage message behavior](https://docs.openclaw.ai/channels/imessage/messaging): Split-send DM coalescing and automatic inbound recovery after a bridge or gateway restart
- [Enabling the imsg private API](https://docs.openclaw.ai/channels/imessage/private-api): Disable SIP and library validation, inject the imsg helper, and verify the private API bridge
- [iMessage private API actions](https://docs.openclaw.ai/channels/imessage/rich-messages): iMessage tapbacks, threaded replies, effects, native polls, approval controls, and question reactions
- [iMessage setup](https://docs.openclaw.ai/channels/imessage/setup): Install the iMessage plugin, set up imsg on the Messages Mac, and grant macOS permissions
- [iMessage troubleshooting](https://docs.openclaw.ai/channels/imessage/troubleshooting): Fixes for a missing imsg binary, silent inbound, ignored DMs and groups, and failed remote attachments
- [IRC](https://docs.openclaw.ai/channels/irc): IRC plugin setup, access controls, and troubleshooting
- [LINE](https://docs.openclaw.ai/channels/line): LINE Messaging API plugin setup, config, and usage
- [Channel location parsing](https://docs.openclaw.ai/channels/location): Channel location parsing and portable outbound location payloads
- [Matrix](https://docs.openclaw.ai/channels/matrix): Matrix support status, setup, and configuration examples
- [Matrix migration](https://docs.openclaw.ai/channels/matrix-migration): How OpenClaw upgrades the previous Matrix plugin in place, including encrypted-state recovery limits and manual recovery steps.
- [Matrix presentation metadata](https://docs.openclaw.ai/channels/matrix-presentation): Matrix MessagePresentation metadata for OpenClaw-aware clients
- [Matrix push rules for quiet previews](https://docs.openclaw.ai/channels/matrix-push-rules): Per-recipient Matrix push rules for quiet finalized preview edits
- [Matrix access control](https://docs.openclaw.ai/channels/matrix/access-control): DM and room policy, allowlists, bot-to-bot traffic, and command authorization
- [Matrix accounts and homeservers](https://docs.openclaw.ai/channels/matrix/accounts-and-homeservers): Multi-account layout, private homeservers, proxies, profiles, and target resolution
- [Matrix encryption and verification](https://docs.openclaw.ai/channels/matrix/encryption): E2EE setup, device verification, cross-signing, and room-key backup
- [Matrix message behavior](https://docs.openclaw.ai/channels/matrix/messaging): Streaming previews, voice-note transcription, and reactions
- [Matrix rich messages and approvals](https://docs.openclaw.ai/channels/matrix/rich-messages): Structured reply controls, approval event metadata, and exec approval prompts
- [Matrix setup](https://docs.openclaw.ai/channels/matrix/setup): Install the plugin, configure a homeserver account, and control which invites the bot accepts
- [Matrix threads and sessions](https://docs.openclaw.ai/channels/matrix/threads-and-sessions): Session routing, reply threading, ACP bindings, and room history context
- [Mattermost](https://docs.openclaw.ai/channels/mattermost): Mattermost bot setup and OpenClaw config
- [Microsoft Teams](https://docs.openclaw.ai/channels/msteams): Microsoft Teams bot support status, capabilities, and configuration
- [Microsoft Teams access control](https://docs.openclaw.ai/channels/msteams/access-control): Microsoft Teams DM and group policy, team and channel allowlists, and conversation IDs
- [Microsoft Teams authentication](https://docs.openclaw.ai/channels/msteams/authentication): Certificate and managed identity authentication for the Microsoft Teams bot
- [Microsoft Teams cards and actions](https://docs.openclaw.ai/channels/msteams/cards-and-actions): Microsoft Teams Adaptive Card approvals, polls, presentation cards, member info, and targets
- [Microsoft Teams configuration](https://docs.openclaw.ai/channels/msteams/configuration): Microsoft Teams configuration keys, environment variables, and history limits
- [Microsoft Teams manifest and permissions](https://docs.openclaw.ai/channels/msteams/manifest-and-permissions): Microsoft Teams app manifest, RSC permissions, and the Graph permissions each capability needs
- [Microsoft Teams message behavior](https://docs.openclaw.ai/channels/msteams/messaging): Microsoft Teams session routing, reply style, attachments, and file sending
- [Microsoft Teams setup](https://docs.openclaw.ai/channels/msteams/setup): Install the Microsoft Teams plugin, register the bot and Teams app, and run against a dev tunnel
- [Microsoft Teams troubleshooting](https://docs.openclaw.ai/channels/msteams/troubleshooting): Microsoft Teams known limitations, common failures, and reference links
- [Nextcloud Talk](https://docs.openclaw.ai/channels/nextcloud-talk): Nextcloud Talk support status, capabilities, and configuration
- [Nostr](https://docs.openclaw.ai/channels/nostr): Nostr DM channel via NIP-04 encrypted messages
- [Pairing](https://docs.openclaw.ai/channels/pairing): Pairing overview: approve who can DM you + which nodes can join
- [QA channel](https://docs.openclaw.ai/channels/qa-channel): Synthetic Slack-class channel plugin for deterministic OpenClaw QA scenarios
- [QQ bot](https://docs.openclaw.ai/channels/qqbot): QQ Bot setup, config, and usage
- [Raft](https://docs.openclaw.ai/channels/raft): Raft External Agent support through the Raft CLI wake bridge
- [Reef](https://docs.openclaw.ai/channels/reef): Reef channel setup: guarded, end-to-end-encrypted messaging between OpenClaw agents of different people
- [Signal](https://docs.openclaw.ai/channels/signal): Signal support via signal-cli (native daemon or bbernhard container), setup paths, and number model
- [Slack](https://docs.openclaw.ai/channels/slack): Slack setup and runtime behavior (Socket Mode, HTTP Request URLs, and relay mode)
- [Slack access control](https://docs.openclaw.ai/channels/slack/access-control): DM policy, channel allowlists, mention gating, and action gates
- [Slack Enterprise Grid](https://docs.openclaw.ai/channels/slack/enterprise-grid): Org-wide Slack installs across every workspace in a Grid organization
- [Slack events and operations](https://docs.openclaw.ai/channels/slack/events): System events, interactions, and optional presence polling
- [Slack manifest and scopes](https://docs.openclaw.ai/channels/slack/manifest-and-scopes): Base app manifest, OAuth scopes, and optional manifest settings
- [Slack media and attachments](https://docs.openclaw.ai/channels/slack/media): Audio clips, inbound file handling, chunking, and delivery targets
- [Slack message behavior](https://docs.openclaw.ai/channels/slack/messaging): Acknowledgement reactions, live streaming previews, and slash commands
- [Slack charts, tables, and approvals](https://docs.openclaw.ai/channels/slack/rich-messages): Native Block Kit charts and tables, modal submissions, and approval buttons
- [Slack setup](https://docs.openclaw.ai/channels/slack/setup): Install the Slack plugin, create the app, and configure tokens
- [Slack threads and sessions](https://docs.openclaw.ai/channels/slack/threads-and-sessions): Session keys, thread replies, reply tags, and Agent View DMs
- [Slack transports](https://docs.openclaw.ai/channels/slack/transports): Socket Mode, HTTP Request URLs, and relay mode compared
- [Slack troubleshooting](https://docs.openclaw.ai/channels/slack/troubleshooting): Fixes for silent channels, ignored DMs, and dead transports
- [SMS](https://docs.openclaw.ai/channels/sms): Twilio SMS/MMS setup, access controls, webhooks, and delivery status
- [Synology Chat](https://docs.openclaw.ai/channels/synology-chat): Synology Chat webhook setup and OpenClaw config
- [Telegram](https://docs.openclaw.ai/channels/telegram): Telegram bot support status, capabilities, and configuration
- [Telegram access control](https://docs.openclaw.ai/channels/telegram/access-control): DM policy, group allowlists, mention gating, and per-chat tool policy
- [Telegram events and operations](https://docs.openclaw.ai/channels/telegram/events): Reaction notifications, config writes from Telegram events, and error reply policy
- [Telegram media and attachments](https://docs.openclaw.ai/channels/telegram/media): Photo albums, voice notes, video notes, locations, venues, and stickers
- [Telegram message behavior](https://docs.openclaw.ai/channels/telegram/messaging): Runtime model, stream previews, native commands, reply tags, ack reactions, and send limits
- [Telegram Control UI Mini App](https://docs.openclaw.ai/channels/telegram/mini-app): Open the OpenClaw Control UI as a Telegram WebApp with /controlui
- [Telegram rich messages and approvals](https://docs.openclaw.ai/channels/telegram/rich-messages): Bot API 10.3 rich messages, inline buttons, message actions, and exec approvals
- [Telegram setup](https://docs.openclaw.ai/channels/telegram/setup): Create the Telegram bot token, configure the channel, and approve the first DM
- [Telegram threads and sessions](https://docs.openclaw.ai/channels/telegram/threads-and-sessions): Forum topic session keys, topic config inheritance, per-topic agents, and ACP bindings
- [Telegram transports](https://docs.openclaw.ai/channels/telegram/transports): Long polling and webhook mode compared, with Gateway routes and durable ingress behavior
- [Telegram troubleshooting](https://docs.openclaw.ai/channels/telegram/troubleshooting): Fixes for silent groups, missing commands, rejected tokens, and unstable polling
- [Tlon](https://docs.openclaw.ai/channels/tlon): Tlon/Urbit support status, capabilities, and configuration
- [Channel troubleshooting](https://docs.openclaw.ai/channels/troubleshooting): Fast channel level troubleshooting with per channel failure signatures and fixes
- [Twitch](https://docs.openclaw.ai/channels/twitch): Twitch chat bot: install, credentials, access control, token refresh
- [WeChat](https://docs.openclaw.ai/channels/wechat): WeChat channel setup through the external openclaw-weixin plugin
- [WeCom](https://docs.openclaw.ai/channels/wecom): Install the external WeCom plugin and find its versioned setup documentation
- [WhatsApp](https://docs.openclaw.ai/channels/whatsapp): WhatsApp channel support, access controls, delivery behavior, and operations
- [X / Twitter](https://docs.openclaw.ai/channels/x): X / Twitter mentions, allowlisted public replies, thread context, and event modes
- [Yuanbao](https://docs.openclaw.ai/channels/yuanbao): Yuanbao bot overview, features, and configuration
- [Zalo](https://docs.openclaw.ai/channels/zalo): Zalo bot support status, capabilities, and configuration
- [Zalo ClawBot](https://docs.openclaw.ai/channels/zaloclawbot): Zalo ClawBot channel setup through the external openclaw-zaloclawbot plugin
- [Zalo personal](https://docs.openclaw.ai/channels/zalouser): Zalo personal account support via native zca-js (QR login), capabilities, and configuration
- [CI pipeline](https://docs.openclaw.ai/ci): CI job graph, scope gates, release umbrellas, and local command equivalents
- [CI capacity and shard weights](https://docs.openclaw.ai/ci/capacity): Runner registration budget, concurrency headroom, and measured shard timings
- [CI checkout ownership](https://docs.openclaw.ai/ci/checkout): Shared checkout anchors, retry budgets, and trusted action policy
- [Local checks and Testbox](https://docs.openclaw.ai/ci/local-proof): Local command equivalents, shrink-only ratchets, and Crabbox remote proof
- [CI pipeline jobs](https://docs.openclaw.ai/ci/pipeline): Job graph, fail-fast order, and the Control UI size budgets
- [Release validation workflows](https://docs.openclaw.ai/ci/release-validation): Full Release Validation, Package Acceptance, install smoke, and Docker E2E
- [Full Release Validation](https://docs.openclaw.ai/ci/release-validation/full-release-validation): The Full Release Validation umbrella, release publish, and Docker Release dispatch
- [Install smoke and Docker E2E](https://docs.openclaw.ai/ci/release-validation/install-smoke-and-docker-e2e): Install Smoke coverage, the local Docker E2E aggregate, and release-path Docker chunks
- [Live and E2E shards](https://docs.openclaw.ai/ci/release-validation/live-and-e2e-shards): Named live and E2E shards in the release live/E2E child workflow
- [Package Acceptance](https://docs.openclaw.ai/ci/release-validation/package-acceptance): Package Acceptance jobs, candidate sources, suite profiles, and dispatch examples
- [Plugin Prerelease](https://docs.openclaw.ai/ci/release-validation/plugin-prerelease): The separate Plugin Prerelease workflow, its batching limits, and when it runs
- [CI routing costs](https://docs.openclaw.ai/ci/routing-costs): Measured CI provider costs, routing decisions, and the 15-minute qualification
- [CI runner classes](https://docs.openclaw.ai/ci/runners): Event-based runner routing, Blacksmith classes, and runner backend modes
- [Scheduled and maintenance workflows](https://docs.openclaw.ai/ci/scheduled-workflows): Performance, QA Lab, CodeQL, Security Review, maintenance jobs, and ClawSweeper forwarding
- [CI scope and routing](https://docs.openclaw.ai/ci/scope-and-routing): Changed-scope detection, lane routing, and manual dispatch behavior
- [Job budgets and platform lanes](https://docs.openclaw.ai/ci/scope-and-routing/job-budgets): UI shards, concurrency and job budgets, lint memory policy, Android rows, and sticky-disk keys
- [Manual dispatches](https://docs.openclaw.ai/ci/scope-and-routing/manual-dispatches): Manual CI dispatch behavior, release-gate fallbacks, and the Windows Testbox Check
- [Node test lanes](https://docs.openclaw.ai/ci/scope-and-routing/node-test-lanes): How the slowest Node test families are split, balanced, packed, and cached
- [Scope selection](https://docs.openclaw.ai/ci/scope-and-routing/selection): Changed-scope detection and the per-area rules that select CI lanes
- [Watch a CI run](https://docs.openclaw.ai/ci/watching-runs): Wait on a pull request head, recover a stuck run, and clear the evidence gate
- [ClawHub](https://docs.openclaw.ai/clawhub): Public ClawHub overview for discovery, install, publish, security, and the clawhub CLI.
- [Acceptable Usage](https://docs.openclaw.ai/clawhub/acceptable-usage): Marketplace policy: what ClawHub allows and what it will not host.
- [API v1](https://docs.openclaw.ai/clawhub/api): Public REST API (v1) overview and conventions.
- [Auth](https://docs.openclaw.ai/clawhub/auth): ClawHub sign-in, API tokens, CLI login, token storage, and revocation.
- [Experimental Claw packages](https://docs.openclaw.ai/clawhub/claws): How to author, publish, and discover experimental Claw packages.
- [CLI](https://docs.openclaw.ai/clawhub/cli): CLI reference: commands, flags, config, and lockfile behavior.
- [Content Rights Requests](https://docs.openclaw.ai/clawhub/content-rights): How to submit a copyright or other content rights concern about material published on ClawHub.
- [How ClawHub Works](https://docs.openclaw.ai/clawhub/how-it-works): How ClawHub listings, versions, installs, publishing, scans, and updates work.
- [HTTP API](https://docs.openclaw.ai/clawhub/http-api): HTTP API reference (public + CLI endpoints + auth).
- [Manage company MCP plugins](https://docs.openclaw.ai/clawhub/managed-mcp): Publish and maintain company MCP integrations through the ClawHub management UI and admin CLI.
- [Moderation and Account Safety](https://docs.openclaw.ai/clawhub/moderation): How ClawHub reports, moderation holds, hidden listings, bans, and account standing work.
- [Org and Namespace Claims](https://docs.openclaw.ai/clawhub/namespace-claims): How to request ClawHub review for org, brand, owner-handle, package-scope, skill-slug, or namespace ownership disputes.
- [Refreshing plugin categories](https://docs.openclaw.ai/clawhub/plugin-category-refresh): Operate the reviewed production plugin category refresh and rollback workflow
- [Plugin validation fixes](https://docs.openclaw.ai/clawhub/plugin-validation-fixes): Fix ClawHub plugin package validation findings before publishing
- [Publishing](https://docs.openclaw.ai/clawhub/publishing): How ClawHub publishing works for skills, plugins, owners, scopes, releases, and review.
- [Quickstart](https://docs.openclaw.ai/clawhub/quickstart): Start using ClawHub: find, install, update, and publish skills or plugins.
- [Security](https://docs.openclaw.ai/clawhub/security): How to report ClawHub security issues and when vulnerabilities are publicly disclosed.
- [Security Audits](https://docs.openclaw.ai/clawhub/security-audits): How to understand ClawHub security audit results before installing a skill or plugin.
- [Skill format](https://docs.openclaw.ai/clawhub/skill-format): Skill folder format, required files, supporting artifacts, limits.
- [Telemetry](https://docs.openclaw.ai/clawhub/telemetry): Install telemetry collected by the ClawHub CLI and how to opt out.
- [Troubleshooting](https://docs.openclaw.ai/clawhub/troubleshooting): Troubleshooting ClawHub sign-in, install, publish, update, and API issues.
- [CLI reference](https://docs.openclaw.ai/cli): OpenClaw CLI index: command list, global flags, and links to per-command pages
- [ACP](https://docs.openclaw.ai/cli/acp): Run the ACP bridge for IDE integrations
- [Agent](https://docs.openclaw.ai/cli/agent): CLI reference for Gateway-backed `openclaw agent` turns and isolated `agent exec` runs
- [Agents](https://docs.openclaw.ai/cli/agents): CLI reference for `openclaw agents` (roles, teams, workspaces, routing, and identity)
- [Approvals](https://docs.openclaw.ai/cli/approvals): CLI reference for `openclaw approvals` and `openclaw exec-policy`
- [Attach CLI](https://docs.openclaw.ai/cli/attach): CLI reference for `openclaw attach` (launch Claude Code with a scoped Gateway MCP grant)
- [Audit records](https://docs.openclaw.ai/cli/audit): CLI reference for activity records, execution identity, and decision receipts
- [Backup](https://docs.openclaw.ai/cli/backup): CLI reference for `openclaw backup` (local and offsite archives, SQLite snapshots, Git history, and schedules)
- [Browser](https://docs.openclaw.ai/cli/browser): CLI reference for `openclaw browser` (lifecycle, profiles, tabs, actions, state, and debugging)
- [Channels](https://docs.openclaw.ai/cli/channels): CLI reference for `openclaw channels` (accounts, status, dead letters, capabilities, resolve, logs, login/logout)
- [Clawbot](https://docs.openclaw.ai/cli/clawbot): CLI reference for `openclaw clawbot` (legacy alias namespace)
- [Claws](https://docs.openclaw.ai/cli/claws): Create, add, update, and remove experimental Claw agent packages
- [Completion](https://docs.openclaw.ai/cli/completion): CLI reference for `openclaw completion` (generate/install shell completion scripts)
- [Config](https://docs.openclaw.ai/cli/config): CLI reference for `openclaw config` (get/set/patch/unset/file/schema/validate)
- [Configure](https://docs.openclaw.ai/cli/configure): CLI reference for `openclaw configure` (interactive configuration prompts)
- [Connect](https://docs.openclaw.ai/cli/connect): Connect a machine to an OpenClaw Gateway with one pasted command
- [Automations (cron)](https://docs.openclaw.ai/cli/cron): CLI reference for `openclaw automations` (schedule and run background jobs)
- [Daemon](https://docs.openclaw.ai/cli/daemon): CLI reference for `openclaw daemon` (legacy alias for gateway service management)
- [Dashboard CLI](https://docs.openclaw.ai/cli/dashboard): CLI reference for `openclaw dashboard` (securely open the Control UI)
- [Devices](https://docs.openclaw.ai/cli/devices): CLI reference for `openclaw devices` (device pairing + token rotation/revocation)
- [Directory](https://docs.openclaw.ai/cli/directory): CLI reference for `openclaw directory` (self, peers, groups)
- [DNS](https://docs.openclaw.ai/cli/dns): CLI reference for `openclaw dns` (wide-area discovery helpers)
- [Docs](https://docs.openclaw.ai/cli/docs): CLI reference for `openclaw docs` (search the live docs index)
- [Doctor CLI](https://docs.openclaw.ai/cli/doctor): CLI reference for `openclaw doctor` (health checks + guided repairs)
- [Other checks and repairs](https://docs.openclaw.ai/cli/doctor/checks): The remaining doctor checks and repairs, from Nix mode to plugins, sandbox, and channels
- [Structured health check contract](https://docs.openclaw.ai/cli/doctor/health-contract): The detect/repair contract that doctor checks and plugin health checks implement
- [Lint and post-upgrade modes](https://docs.openclaw.ai/cli/doctor/lint): Read-only lint findings, check selection, and post-upgrade plugin checks
- [Gateway and service recovery](https://docs.openclaw.ai/cli/doctor/recovery): Recover the Gateway service, a remote Gateway, Control UI assets, and Gateway tokens
- [Run doctor](https://docs.openclaw.ai/cli/doctor/running): Doctor postures, example invocations, and the full option table
- [SQLite maintenance and session migration](https://docs.openclaw.ai/cli/doctor/sqlite-maintenance): Shared-state compaction plus targeted session SQLite inspection, import, and recovery
- [Legacy state migration](https://docs.openclaw.ai/cli/doctor/state-migrations): How `doctor --fix` migrates legacy file-backed state into SQLite
- [File transfers](https://docs.openclaw.ai/cli/file-transfer): CLI reference for `openclaw file-transfer` (review and migrate standing file-transfer approvals)
- [Gateway](https://docs.openclaw.ai/cli/gateway): OpenClaw Gateway CLI (`openclaw gateway`) — run, query, and discover gateways
- [Discover gateways (Bonjour)](https://docs.openclaw.ai/cli/gateway/discovery): `openclaw gateway discover` and the Bonjour beacons and TXT hints it scans for
- [Query a running Gateway](https://docs.openclaw.ai/cli/gateway/query): Query a running Gateway: health, usage-cost, stability, diagnostics export, status, connectivity checks, call, suspend, and resume
- [Restart and supervision](https://docs.openclaw.ai/cli/gateway/restart-and-supervision): `openclaw gateway restart`, install identity, external process supervisors, and Gateway profiling
- [Run the Gateway](https://docs.openclaw.ai/cli/gateway/running): `openclaw gateway` run options, startup behavior, and revealing the configured token
- [Manage the Gateway service](https://docs.openclaw.ai/cli/gateway/service): Install, start, stop, and uninstall the managed Gateway service, including wrappers, heap sizing, and install-time auth
- [Health](https://docs.openclaw.ai/cli/health): CLI reference for `openclaw health` (gateway health snapshot via RPC)
- [Hooks CLI](https://docs.openclaw.ai/cli/hooks): CLI reference for internal hook discovery, eligibility, enablement, and hook packs
- [Inference CLI](https://docs.openclaw.ai/cli/infer): Infer-first CLI for provider-backed model, image, audio, TTS, video, web, and embedding workflows
- [Logs](https://docs.openclaw.ai/cli/logs): CLI reference for `openclaw logs` (tail gateway logs via RPC)
- [MCP](https://docs.openclaw.ai/cli/mcp): Expose OpenClaw channel conversations over MCP and manage saved MCP server definitions
- [MCP Apps](https://docs.openclaw.ai/cli/mcp/apps): Enable and secure the MCP Apps host bridge that renders server-provided HTML views
- [MCP in the Control UI](https://docs.openclaw.ai/cli/mcp/control-ui): Edit and inspect MCP servers from the browser Control UI settings page
- [JSON output shapes](https://docs.openclaw.ai/cli/mcp/json-output): JSON output shapes for openclaw mcp status, doctor, and check
- [Manage saved MCP servers](https://docs.openclaw.ai/cli/mcp/registry): Manage OpenClaw-saved MCP server definitions with the mcp registry subcommands
- [Run OpenClaw as an MCP server](https://docs.openclaw.ai/cli/mcp/serve): Run OpenClaw as a stdio MCP server so an MCP client can read and send channel conversations
- [Transports and OAuth](https://docs.openclaw.ai/cli/mcp/transports): Stdio, SSE/HTTP, and Streamable HTTP transport fields plus the MCP OAuth workflow
- [Memory](https://docs.openclaw.ai/cli/memory): CLI reference for `openclaw memory` (status/index/reset/search/forget/promote/promote-explain/rem-harness/rem-backfill/session-backfill)
- [Message](https://docs.openclaw.ai/cli/message): CLI reference for `openclaw message` (send + channel actions)
- [Migrate](https://docs.openclaw.ai/cli/migrate): CLI reference for `openclaw migrate` (import state from another agent system)
- [Models](https://docs.openclaw.ai/cli/models): CLI reference for `openclaw models` (status/list/set/scan, aliases, fallbacks, shared auth, personal accounts)
- [Node](https://docs.openclaw.ai/cli/node): CLI reference for `openclaw node` (headless node host)
- [Nodes CLI](https://docs.openclaw.ai/cli/nodes): CLI reference for `openclaw nodes` (status, pairing, invoke, camera/screen/location/notify and the macOS widget panel)
- [Onboard](https://docs.openclaw.ai/cli/onboard): CLI reference for `openclaw onboard` (interactive onboarding)
- [OpenClaw setup agent](https://docs.openclaw.ai/cli/openclaw): CLI reference and security model for the inference-backed OpenClaw setup and repair helper
- [Pairing CLI](https://docs.openclaw.ai/cli/pairing): CLI reference for `openclaw pairing` (approve/list pairing requests)
- [Path](https://docs.openclaw.ai/cli/path): CLI reference for `openclaw path` (inspect and edit workspace files via the `oc://` addressing scheme)
- [Plugins CLI](https://docs.openclaw.ai/cli/plugins): CLI reference for `openclaw plugins` (init, build, validate, list, install, reload, marketplace, uninstall, enable/disable, doctor)
- [Author plugins](https://docs.openclaw.ai/cli/plugins/authoring): Scaffold, build, validate, and pack an OpenClaw plugin with `openclaw plugins init`
- [Inspect and diagnose plugins](https://docs.openclaw.ai/cli/plugins/inspect-and-diagnose): `plugins inspect`, `plugins doctor`, and `plugins registry` for plugin state, load failures, and the cold registry
- [Install plugins](https://docs.openclaw.ai/cli/plugins/install): Install sources and locators, trust and install policy, capability consent, and marketplace installs
- [List installed plugins](https://docs.openclaw.ai/cli/plugins/list): `plugins list` output and options, discovery diagnostics, and the persisted plugin index
- [Marketplace feeds](https://docs.openclaw.ai/cli/plugins/marketplace): `plugins marketplace` entries, list, and refresh, plus hosted feed trust and signed feed profiles
- [Uninstall and update plugins](https://docs.openclaw.ai/cli/plugins/uninstall-and-update): What `plugins uninstall` removes, how `plugins update` resolves sources, channels, pins, and integrity drift, and reloading edited plugins
- [Policy](https://docs.openclaw.ai/cli/policy): CLI reference for `openclaw policy` conformance checks
- [Accept and watch policy state](https://docs.openclaw.ai/cli/policy/attestation): Policy evidence, the attestation hash tuple, and `policy watch` drift detection
- [Author a policy file](https://docs.openclaw.ai/cli/policy/authoring): Enable the Policy plugin and author `policy.jsonc`, with a minimal example covering every supported section
- [Policy findings, repair, and exit codes](https://docs.openclaw.ai/cli/policy/findings): Every policy check id, what `doctor --fix` will and will not repair, and command exit codes
- [Policy rule reference](https://docs.openclaw.ai/cli/policy/rules): Every supported `policy.jsonc` rule, the OpenClaw state it observes, and when to use it
- [Run and configure policy checks](https://docs.openclaw.ai/cli/policy/running-checks): `policy check`, `policy compare`, and the `plugins.entries.policy.config` settings that govern them
- [Scoped policy overlays](https://docs.openclaw.ai/cli/policy/scopes): Use `scopes.<scopeName>` to hold specific agents or channels to stricter policy than the baseline
- [Promos](https://docs.openclaw.ai/cli/promos): CLI reference for `openclaw promos` (list and claim promotional model offers)
- [Proxy](https://docs.openclaw.ai/cli/proxy): CLI reference for `openclaw proxy`, including operator-managed proxy validation and the local debug proxy capture inspector
- [QR](https://docs.openclaw.ai/cli/qr): CLI reference for `openclaw qr` (generate mobile pairing QR + setup code)
- [Reset](https://docs.openclaw.ai/cli/reset): CLI reference for `openclaw reset` (reset local state/config)
- [Resume](https://docs.openclaw.ai/cli/resume): CLI reference for attaching the TUI to a recent Gateway session
- [Sandbox CLI](https://docs.openclaw.ai/cli/sandbox): Manage sandbox runtimes and inspect effective sandbox policy
- [Secrets CLI](https://docs.openclaw.ai/cli/secrets): CLI reference for `openclaw secrets` (store, reload, audit, configure, apply)
- [Security CLI](https://docs.openclaw.ai/cli/security): CLI reference for `openclaw security` (audit and fix common security footguns)
- [Sessions](https://docs.openclaw.ai/cli/sessions): CLI reference for listing, importing, archiving, deleting, and maintaining stored sessions
- [Setup CLI](https://docs.openclaw.ai/cli/setup): CLI reference for `openclaw setup` (system-agent chat with onboarding fallback)
- [Skills CLI](https://docs.openclaw.ai/cli/skills): CLI reference for `openclaw skills` (search/install/update/verify/list/info/check/library/workshop)
- [openclaw status](https://docs.openclaw.ai/cli/status): CLI reference for `openclaw status` (diagnostics, checks, usage snapshots)
- [storage](https://docs.openclaw.ai/cli/storage): List, initialize, and test configured storage locations
- [System](https://docs.openclaw.ai/cli/system): CLI reference for `openclaw system` (system events, heartbeat, presence)
- [Transcripts CLI](https://docs.openclaw.ai/cli/transcripts): CLI reference for `openclaw transcripts` (list, show, and export stored transcripts)
- [Triage](https://docs.openclaw.ai/cli/triage): CLI reference for `openclaw triage` (sanitized diagnostics and agent handoff)
- [openclaw tui](https://docs.openclaw.ai/cli/tui): CLI reference for `openclaw tui` (Gateway-backed or local embedded terminal UI)
- [Uninstall CLI](https://docs.openclaw.ai/cli/uninstall): CLI reference for `openclaw uninstall` (remove gateway service + local data)
- [Update](https://docs.openclaw.ai/cli/update): CLI reference for `openclaw update` (updates, repair, and recovery cleanup)
- [How an update runs](https://docs.openclaw.ai/cli/update/how-updates-run): How `openclaw update` switches channels, checks the new version, hands off the restart, and updates a Git checkout
- [Update repair and recovery](https://docs.openclaw.ai/cli/update/repair-and-recovery): Recovering from a failed `openclaw update`, plus the `update repair` and `update cleanup` subcommands
- [Update status and run history](https://docs.openclaw.ai/cli/update/status-and-history): `openclaw update status` plus the durable run ledger, reports, and artifacts every update writes
- [Users](https://docs.openclaw.ai/cli/users): CLI reference for `openclaw users` (profiles, email aliases, and duplicate merges)
- [Voicecall](https://docs.openclaw.ai/cli/voicecall): CLI reference for `openclaw voicecall` (voice-call plugin command surface)
- [Webhooks](https://docs.openclaw.ai/cli/webhooks): CLI reference for `openclaw webhooks` (Gmail Pub/Sub setup and runner)
- [Wiki](https://docs.openclaw.ai/cli/wiki): CLI reference for `openclaw wiki` (memory-wiki vault status, search, compile, lint, apply, bridge, ChatGPT import, and Obsidian helpers)
- [Workboard CLI](https://docs.openclaw.ai/cli/workboard): CLI reference for `openclaw workboard` cards, dispatch, and worker runs
- [Worker](https://docs.openclaw.ai/cli/worker): Internal operator reference for the restricted cloud worker runtime
- [Active memory](https://docs.openclaw.ai/concepts/active-memory): Deep conversation-history recall that escalates only when deterministic memory recall is insufficient
- [Advanced overrides and transcripts](https://docs.openclaw.ai/concepts/active-memory/advanced-options): Escape hatches for thinking, fast mode, and the sub-agent prompt, plus exporting blocking sub-agent transcripts for debugging.
- [Configuration reference](https://docs.openclaw.ai/concepts/active-memory/configuration): Every plugins.entries.active-memory key: type, meaning, range, and default, plus the extra tuning fields.
- [Enabling active memory](https://docs.openclaw.ai/concepts/active-memory/enabling): Two ways to turn active memory on: the per-agent rememberAcrossConversations product setting, and the advanced plugin quick start.
- [How active memory works](https://docs.openclaw.ai/concepts/active-memory/how-it-works): The escalation flow, which surfaces run active memory, the two targeting paths, and the session types the advanced path allows.
- [Memory tools](https://docs.openclaw.ai/concepts/active-memory/memory-tools): The concrete recall tools the blocking sub-agent may call, and per-provider setups for built-in memory, LanceDB, and Lossless Claw.
- [Recommended setup](https://docs.openclaw.ai/concepts/active-memory/recommended-setup): The starter configuration to begin with, and the cold-start grace budget to set when upgrading from v2026.4.x.
- [Session controls](https://docs.openclaw.ai/concepts/active-memory/session-controls): Pause or resume active memory per session or globally, and surface its status, debug summary, and raw hidden prefix in a chat.
- [Troubleshooting active memory](https://docs.openclaw.ai/concepts/active-memory/troubleshooting): A checklist for when active memory does not run, and the common recall failures with their causes.
- [Query modes, prompts, and models](https://docs.openclaw.ai/concepts/active-memory/tuning): How much conversation the sub-agent sees, how eager it is about returning memory, and how its model is resolved.
- [Agent runtime](https://docs.openclaw.ai/concepts/agent): Agent runtime, workspace contract, and session bootstrap
- [Agent bindings](https://docs.openclaw.ai/concepts/agent-bindings): Route channel accounts and conversations to the right OpenClaw agent
- [Agent loop](https://docs.openclaw.ai/concepts/agent-loop): Agent loop lifecycle, streams, and wait semantics
- [Agent runtimes](https://docs.openclaw.ai/concepts/agent-runtimes): How OpenClaw separates model providers, models, channels, and agent runtimes
- [Agent workspace](https://docs.openclaw.ai/concepts/agent-workspace): Agent workspace: location, layout, and backup strategy
- [Gateway architecture](https://docs.openclaw.ai/concepts/architecture): WebSocket gateway architecture, components, and client flows
- [Compaction](https://docs.openclaw.ai/concepts/compaction): How OpenClaw summarizes long conversations to stay within model limits
- [Context](https://docs.openclaw.ai/concepts/context): Context: what the model sees, how it is built, and how to inspect it
- [Context engine](https://docs.openclaw.ai/concepts/context-engine): Context engine: pluggable context assembly, compaction, and subagent lifecycle
- [Decision models](https://docs.openclaw.ai/concepts/decision-models): Choose decision models and define typed choices, scores, and Boolean judgments
- [Delegate architecture](https://docs.openclaw.ai/concepts/delegate-architecture): Delegate architecture: running OpenClaw as a named agent on behalf of an organization
- [Dreaming](https://docs.openclaw.ai/concepts/dreaming): Background memory consolidation with light, deep, and REM phases plus a Dream Diary
- [Experimental features](https://docs.openclaw.ai/concepts/experimental-features): What experimental flags mean in OpenClaw and which ones are currently documented
- [Features](https://docs.openclaw.ai/concepts/features): OpenClaw capabilities across channels, routing, media, and UX.
- [The main session](https://docs.openclaw.ai/concepts/main-session): One rolling conversation across all your channels: the personal-agent default
- [Managed worktrees](https://docs.openclaw.ai/concepts/managed-worktrees): Run agent tasks in isolated git checkouts with automatic snapshots and cleanup
- [Mantis](https://docs.openclaw.ai/concepts/mantis): Mantis captures visual end-to-end evidence for live transport comparisons and focused candidate-only browser proofs, then attaches the artifacts to PRs.
- [Mantis Slack desktop runbook](https://docs.openclaw.ai/concepts/mantis-slack-desktop-runbook): Operator runbook for Mantis Slack desktop QA: GitHub dispatch, local CLI, warm VNC leases, hydrate modes, timing interpretation, artifacts, and failure handling.
- [Markdown formatting](https://docs.openclaw.ai/concepts/markdown-formatting): Markdown formatting pipeline for outbound channels
- [Memory overview](https://docs.openclaw.ai/concepts/memory): How OpenClaw remembers things across sessions
- [Memory architecture](https://docs.openclaw.ai/concepts/memory-architecture): End-to-end architecture of OpenClaw memory: tiers, provenance, dreaming, recall lanes, the user model, and standing intents
- [Builtin memory engine](https://docs.openclaw.ai/concepts/memory-builtin): The default SQLite-based memory backend with keyword, vector, and hybrid search
- [Honcho memory](https://docs.openclaw.ai/concepts/memory-honcho): AI-native cross-session memory via the Honcho plugin
- [Memory provenance and deletion](https://docs.openclaw.ai/concepts/memory-provenance): Trace session-derived memories, control ingestion, and preview or delete tracked memory artifacts
- [Memory search](https://docs.openclaw.ai/concepts/memory-search): How memory search finds relevant notes using embeddings and hybrid retrieval
- [Messages](https://docs.openclaw.ai/concepts/messages): Message flow, sessions, queueing, and reasoning visibility
- [Model failover](https://docs.openclaw.ai/concepts/model-failover): How OpenClaw rotates auth profiles and falls back across models
- [Model providers](https://docs.openclaw.ai/concepts/model-providers): Index of the model provider reference: quick rules, Control UI and keys, bundled provider plugins, and custom providers
- [Control UI and API keys](https://docs.openclaw.ai/concepts/model-providers/control-ui-and-keys): Configuring providers from Settings -> Models, plugin-owned provider behavior, and API key rotation.
- [Custom providers and local runtimes](https://docs.openclaw.ai/concepts/model-providers/custom-providers): Providers configured through models.providers: custom endpoints, base URLs, and local inference servers.
- [Official provider plugins](https://docs.openclaw.ai/concepts/model-providers/official-provider-plugins): Per-provider setup for the bundled provider plugins, the bundled provider table, and provider quirks.
- [Quick rules](https://docs.openclaw.ai/concepts/model-providers/quick-rules): Model refs, CLI helpers, the primary-model preservation rule, and the OpenAI provider/runtime split.
- [Models CLI](https://docs.openclaw.ai/concepts/models): How OpenClaw resolves provider/model refs, config keys, and the `/model` chat command
- [Multi-agent routing](https://docs.openclaw.ai/concepts/multi-agent): Multi-agent routing: agent boundaries, channel accounts, and bindings
- [Multi-user mode](https://docs.openclaw.ai/concepts/multi-user): How session ownership, presence, and human mentions work when several people operate one agent
- [OAuth](https://docs.openclaw.ai/concepts/oauth): OAuth in OpenClaw: token exchange, storage, and multi-account patterns
- [Parallel specialist lanes](https://docs.openclaw.ai/concepts/parallel-specialist-lanes): Run parallel specialist agents without clogging shared model and tool capacity
- [Personal agent benchmark pack](https://docs.openclaw.ai/concepts/personal-agent-benchmark-pack): Local qa-channel scenarios for privacy-preserving personal assistant workflow checks.
- [Presence](https://docs.openclaw.ai/concepts/presence): How OpenClaw presence entries are produced, merged, and displayed
- [Progress drafts](https://docs.openclaw.ai/concepts/progress-drafts): Progress drafts: one visible work-in-progress message that updates while an agent runs
- [QA overview](https://docs.openclaw.ai/concepts/qa-e2e-automation): Index of the private QA stack: qa-lab, qa-channel, repo-backed scenarios, live transport lanes, transport adapters, and reporting.
- [Channel QA reference](https://docs.openclaw.ai/concepts/qa-e2e-automation/channel-qa-reference): Shared CLI flags for the real-transport lanes and the Buzz, Telegram, and Discord lane reference.
- [Command surface](https://docs.openclaw.ai/concepts/qa-e2e-automation/command-surface): The `pnpm openclaw qa` subcommand table and the profile-backed `qa run` selector.
- [Extending the QA stack](https://docs.openclaw.ai/concepts/qa-e2e-automation/extending-the-stack): Repo-backed seed assets, provider mock lanes, transport adapters, and what adding a channel takes.
- [Operator flow](https://docs.openclaw.ai/concepts/qa-e2e-automation/operator-flow): Bring up QA Lab and run the observability, Matrix, Discord Mantis, Slack desktop, and credential-pool lanes.
- [QA reporting](https://docs.openclaw.ai/concepts/qa-e2e-automation/qa-reporting): QA report and evidence artifacts, the character-eval judged report, and where each run writes them.
- [Canonical scenario coverage](https://docs.openclaw.ai/concepts/qa-e2e-automation/scenario-coverage): How taxonomy coverage IDs, scenario YAML, and profiles select what a QA run executes, plus the Multipass suite runner.
- [Slack QA](https://docs.openclaw.ai/concepts/qa-e2e-automation/slack-qa): The Slack QA lane and the workspace, app, and scope provisioning it needs.
- [WhatsApp QA and credentials](https://docs.openclaw.ai/concepts/qa-e2e-automation/whatsapp-and-credentials): The WhatsApp QA lane and the shared Convex credential pool used by the real-transport lanes.
- [Command queue](https://docs.openclaw.ai/concepts/queue): Auto-reply queue modes, shared background capacity, and per-session overrides
- [Steering queue](https://docs.openclaw.ai/concepts/queue-steering): How active-run steering queues messages at runtime boundaries
- [Recommended models](https://docs.openclaw.ai/concepts/recommended-models): How the curated recommended model list is reviewed, ordered, and published
- [Retry policy](https://docs.openclaw.ai/concepts/retry): Retry policy for outbound provider calls
- [Session management](https://docs.openclaw.ai/concepts/session): How OpenClaw manages conversation sessions
- [Session synchronization and attachment](https://docs.openclaw.ai/concepts/session-attachment): How Gateway-owned sessions continue across the Control UI, terminal, CLI, mobile clients, and coding harnesses
- [Session pruning](https://docs.openclaw.ai/concepts/session-pruning): Trimming old tool results to keep context lean and caching efficient
- [Session search](https://docs.openclaw.ai/concepts/session-search): Search past session transcripts and reopen the matching context
- [Session state awareness](https://docs.openclaw.ai/concepts/session-state): Durable session state signal log: state versions, watchers, stale-state notices, and reconciliation
- [Session tools](https://docs.openclaw.ai/concepts/session-tool): Agent tools for cross-session status, recall, messaging, and sub-agent orchestration
- [SOUL.md personality guide](https://docs.openclaw.ai/concepts/soul): Use SOUL.md to give your OpenClaw agent an actual voice instead of generic assistant sludge
- [Standing intents](https://docs.openclaw.ai/concepts/standing-intents): Remember event-conditioned future actions without relying on long conversational context
- [Storage locations](https://docs.openclaw.ai/concepts/storage-locations): Named storage destinations, explicit initialization, encryption, and provider configuration
- [Streaming and chunking](https://docs.openclaw.ai/concepts/streaming): Streaming + chunking behavior (block replies, channel preview streaming, mode mapping)
- [Subagent yield handoff](https://docs.openclaw.ai/concepts/subagent-yield-handoff): Ownership of subagent completion and progress when a requester yields
- [System prompt](https://docs.openclaw.ai/concepts/system-prompt): What the OpenClaw system prompt contains and how it is assembled
- [Timezones](https://docs.openclaw.ai/concepts/timezone): Where timezones show up in OpenClaw — envelopes, tool payloads, system prompt
- [TypeBox](https://docs.openclaw.ai/concepts/typebox): TypeBox schemas as the single source of truth for the gateway protocol
- [Typing indicators](https://docs.openclaw.ai/concepts/typing-indicators): When OpenClaw shows typing indicators and how to tune them
- [Usage tracking](https://docs.openclaw.ai/concepts/usage-tracking): Usage tracking surfaces and credential requirements
- [User model](https://docs.openclaw.ai/concepts/user-model): Manage durable user preferences and your Gateway profile identity
- [Date and time](https://docs.openclaw.ai/date-time): Date and time handling across envelopes, prompts, tools, and connectors
- [Diagnostics flags](https://docs.openclaw.ai/diagnostics/flags): Diagnostics flags for targeted debug logs
- [Docs map](https://docs.openclaw.ai/docs_map): Generated heading map for OpenClaw docs pages
- [Gateway runbook](https://docs.openclaw.ai/gateway): Runbook for the Gateway service, lifecycle, and operations
- [1Password](https://docs.openclaw.ai/gateway/1password): Use the 1Password plugin, bundled skill, or official MCP with OpenClaw
- [Audit history](https://docs.openclaw.ai/gateway/audit): Metadata-only activity history plus durable run identity and decision receipts
- [Authentication](https://docs.openclaw.ai/gateway/authentication): Model authentication: OAuth, API keys, Claude CLI reuse, and Anthropic setup-token
- [Background exec and process tool](https://docs.openclaw.ai/gateway/background-process): Background exec execution and process management
- [Bonjour discovery](https://docs.openclaw.ai/gateway/bonjour): Bonjour/mDNS discovery + debugging (Gateway beacons, clients, and common failure modes)
- [CLI backends](https://docs.openclaw.ai/gateway/cli-backends): CLI backends: run agent turns through a local AI CLI, with an optional MCP tool bridge
- [Building a Gateway client](https://docs.openclaw.ai/gateway/clients): Build a third-party operator or WebChat client for the Gateway WebSocket protocol
- [Cloud Sessions](https://docs.openclaw.ai/gateway/cloud-sessions): Run sessions on paired devices or throwaway cloud machines while the Gateway keeps the transcript, workspace, and credentials
- [Cloud Workers](https://docs.openclaw.ai/gateway/cloud-workers): Dispatch session work to throwaway cloud machines with OpenClaw worker turns or Codex remote execution
- [Cloud Worker Desktop](https://docs.openclaw.ai/gateway/cloud-workers/desktop): Watch or control a desktop-capable cloud worker from the Control UI
- [Dispatching a cloud session](https://docs.openclaw.ai/gateway/cloud-workers/dispatching-a-session): Eligibility gates, the Control UI Place picker, cloud child sessions, and runtime support
- [Worker-local inference](https://docs.openclaw.ai/gateway/cloud-workers/native-inference): Worker inference on an externally managed paired worker host
- [Per-project default profiles](https://docs.openclaw.ai/gateway/cloud-workers/per-project-default-profiles): Map a repository to a default cloud worker profile, and how provision replay adopts a fixed lease
- [Placement and machine selection](https://docs.openclaw.ai/gateway/cloud-workers/placement-and-machine-selection): Codex on a paired device, either harness on a cloud profile, and per-session operating system and machine class
- [Cloud worker security model](https://docs.openclaw.ai/gateway/cloud-workers/security-model): Worker ingress, tool authority, minted credentials, enrollment binding, and credential boundaries
- [Cloud session lifecycle and durability](https://docs.openclaw.ai/gateway/cloud-workers/session-lifecycle): Dispatch, workspace reconciliation, moves, stop and reclaim, recovery, and what survives a dead machine
- [Worker setup and bundle installation](https://docs.openclaw.ai/gateway/cloud-workers/setup-and-bundle-installation): The idempotent setup command, Gateway-prepared runtime archives, and building a custom node package
- [Cloud worker troubleshooting](https://docs.openclaw.ai/gateway/cloud-workers/troubleshooting): Symptoms and fixes for profile advertisement, authorization, bootstrap, enrollment, and teardown
- [Verify a cloud worker profile](https://docs.openclaw.ai/gateway/cloud-workers/verify-the-profile): Validate config, confirm the profile is advertised, and prove a new cloud worker profile end to end
- [Cloud worker warm images](https://docs.openclaw.ai/gateway/cloud-workers/warm-images): Capture, reuse, refresh, and recover prepared cloud-worker images
- [Cloudflare Tunnel and Access](https://docs.openclaw.ai/gateway/cloudflare-access): Publish a loopback Gateway through a Cloudflare Tunnel and authenticate every client with Cloudflare Access
- [Configuration — agents](https://docs.openclaw.ai/gateway/config-agents): Agent defaults, multi-agent routing, session, messages, and talk config
- [Configuration — per-agent entries and multi-agent routing](https://docs.openclaw.ai/gateway/config-agents/entries-and-multi-agent): agents.entries overrides, multiAgent bindings, binding match fields, and access profiles
- [Configuration — agent heartbeat, compaction, and streaming](https://docs.openclaw.ai/gateway/config-agents/heartbeat-compaction-and-streaming): Heartbeat runs, system agent, compaction, context pruning, block streaming, and typing indicators
- [Configuration — messages and talk](https://docs.openclaw.ai/gateway/config-agents/messages-and-talk): messages.* delivery, prefixes, ack reactions, queueing, TTS, and talk.* defaults
- [Configuration — agent models](https://docs.openclaw.ai/gateway/config-agents/models): Default model, fallback chain, per-purpose model slots, and model selection scope
- [Configuration — agent runtime and CLI backends](https://docs.openclaw.ai/gateway/config-agents/runtime-and-cli-backends): Agent runtime policy on providers and models, CLI backend selection, and GPT-5 personality
- [Configuration — agent sandboxing](https://docs.openclaw.ai/gateway/config-agents/sandbox): The agents.defaults.sandbox block: image, mounts, workspace mode, and network policy
- [Configuration — agent sessions](https://docs.openclaw.ai/gateway/config-agents/sessions): session.* scope, identity links, reset policy, sharing, and retention
- [Configuration — agent workspace and bootstrap](https://docs.openclaw.ai/gateway/config-agents/workspace-and-bootstrap): Agent workspace paths, bootstrap injection, context budgets, image handling, and timezone
- [Configuration — automations and media template variables](https://docs.openclaw.ai/gateway/config-automation): Automation config under cron plus the media model template variable surface
- [Configuration — browser, UI, and desktop](https://docs.openclaw.ai/gateway/config-browser-ui-desktop): Browser automation, Control UI presentation, and desktop or paired-node config
- [Configuration — channels](https://docs.openclaw.ai/gateway/config-channels): Channel configuration: access control, pairing, per-channel keys across Slack, Discord, Telegram, WhatsApp, Matrix, iMessage, and more
- [Configuration — chat commands](https://docs.openclaw.ai/gateway/config-channels/commands): The commands.* block: native and text command surfaces, bash, config writes, and owner allowlists
- [Configuration — community chat channels](https://docs.openclaw.ai/gateway/config-channels/community-chat): Per-channel config keys for Discord, Matrix, and IRC
- [Configuration — group mention gating and history](https://docs.openclaw.ai/gateway/config-channels/mention-gating-and-history): Group mention gating, visible reply modes, DM history limits, and self-chat mode
- [Configuration — personal messaging channels](https://docs.openclaw.ai/gateway/config-channels/personal-messaging): Per-channel config keys for WhatsApp, Telegram, Signal, iMessage, and LINE
- [Configuration — shared channel policies](https://docs.openclaw.ai/gateway/config-channels/shared-policies): Shared channel keys: DM and group access policies, model overrides, channel defaults, and multi-account setup
- [Configuration — workplace chat channels](https://docs.openclaw.ai/gateway/config-channels/workplace-chat): Per-channel config keys for Google Chat, Slack, Mattermost, and Microsoft Teams
- [Configuration — cloud worker environments](https://docs.openclaw.ai/gateway/config-cloud-workers): Cloud worker profiles under cloudWorkers, including Crabbox and static SSH development
- [Configuration — MCP, skills, and plugins](https://docs.openclaw.ai/gateway/config-extensions): Extension config: MCP servers, skills, plugin entries, and the canvas widget presenter
- [Configuration — gateway](https://docs.openclaw.ai/gateway/config-gateway): Gateway config: bind, auth, roles, Control UI, terminal, remote, nodes, TLS, and reload
- [Configuration — hooks](https://docs.openclaw.ai/gateway/config-hooks): Hook config: HTTP contract, agent payload, session policy, mapping, retries, and Gmail
- [Configuration — audit, logging, diagnostics, and telemetry](https://docs.openclaw.ai/gateway/config-observability): Observability config: audit, logging, diagnostics, and telemetry keys
- [Configuration — runtime basics](https://docs.openclaw.ai/gateway/config-runtime): Runtime config: worktree storage, capacity, and acceleration, model routing, discovery, updates, ACP, and the wizard
- [Configuration — environment, secrets, and includes](https://docs.openclaw.ai/gateway/config-secrets-env): Environment variables, secret providers, auth storage, and $include config splitting
- [Configuration — tools and custom providers](https://docs.openclaw.ai/gateway/config-tools): Tools config (policy, experimental toggles, provider-backed tools) and custom provider/base-URL setup
- [Configuration — built-in tool settings](https://docs.openclaw.ai/gateway/config-tools/built-in-tools): Per-tool settings for exec, loop detection, web search and fetch, media understanding, and the plan tool
- [Configuration — custom providers and base URLs](https://docs.openclaw.ai/gateway/config-tools/custom-providers): models.providers registration, base-URL trust, and the full provider field reference
- [Configuration — GitHub identity for agent tools](https://docs.openclaw.ai/gateway/config-tools/github-identity): tools.github: the shared managed GitHub CLI identity, its refresh, and its execution boundaries
- [Configuration — provider examples](https://docs.openclaw.ai/gateway/config-tools/provider-examples): Worked custom-provider configurations for Cerebras, Kimi, llama.cpp, LM Studio, MiniMax, Moonshot, OpenCode, Synthetic, and Z.AI
- [Configuration — cross-agent, session, and subagent tools](https://docs.openclaw.ai/gateway/config-tools/sessions-and-subagents): tools.agentToAgent, tools.sessions visibility, sessions_spawn attachments, and subagent defaults
- [Configuration — tool policy](https://docs.openclaw.ai/gateway/config-tools/tool-policy): Tool profiles, groups, sandbox tool gates, code mode, and allow/deny policy layers
- [Configuration](https://docs.openclaw.ai/gateway/configuration): Configuration overview: common tasks, quick setup, and links to the Configuration reference
- [Configuration examples](https://docs.openclaw.ai/gateway/configuration-examples): Schema-accurate configuration examples for common OpenClaw setups
- [Configuration reference](https://docs.openclaw.ai/gateway/configuration-reference): Gateway config reference for core OpenClaw keys, defaults, and links to dedicated subsystem references
- [Common tasks](https://docs.openclaw.ai/gateway/configuration/common-tasks): Task-oriented configuration recipes: channels, models, access, sessions, sandboxing, cron, hooks, routing, and $include
- [Config RPC (programmatic updates)](https://docs.openclaw.ai/gateway/configuration/config-rpc): Writing config over the gateway API: config.get, config.patch, config.apply, and replacePaths
- [Environment variables](https://docs.openclaw.ai/gateway/configuration/environment-variables): Env var sources and precedence, inline env config, shell env import, substitution, and secret refs
- [Config hot reload](https://docs.openclaw.ai/gateway/configuration/hot-reload): How the Gateway applies config snapshots, which changes hot-apply, and which need a restart
- [Diagnostics export](https://docs.openclaw.ai/gateway/diagnostics): Create shareable Gateway diagnostics bundles for bug reports
- [Discovery and transports](https://docs.openclaw.ai/gateway/discovery): Node discovery and transports (Bonjour, Tailscale, SSH) for finding the Gateway
- [Doctor](https://docs.openclaw.ai/gateway/doctor): Doctor command: health checks, config migrations, and repair steps
- [What doctor checks](https://docs.openclaw.ai/gateway/doctor/checks): Summary of every repair, migration, and health check doctor runs
- [Config and migration repairs](https://docs.openclaw.ai/gateway/doctor/config-migrations): Config normalization, legacy config key migrations, and update-time schema publication
- [Gateway, service, and security checks](https://docs.openclaw.ai/gateway/doctor/gateway-and-services): Gateway service, pairing, security, workspace status, auth, health, and supervisor checks (checks 8-17)
- [Provider and route repairs](https://docs.openclaw.ai/gateway/doctor/provider-repairs): Provider override, browser, OAuth TLS, and route repairs (checks 2b-2g)
- [Run doctor](https://docs.openclaw.ai/gateway/doctor/running): Run doctor, pick a mode, and read the read-only lint report
- [State, session, and plugin repairs](https://docs.openclaw.ai/gateway/doctor/state-and-sessions): Disk layout, cron store, session, model auth, sandbox, and plugin install repairs (checks 3-7b)
- [Workspace tips and Dreams UI actions](https://docs.openclaw.ai/gateway/doctor/workspace-and-dreams): Config write, workspace tips, repointed aliases, and the Control UI Dreams actions
- [Embedding OpenClaw](https://docs.openclaw.ai/gateway/embedding): Supervise the OpenClaw Gateway as a child process from Electron or another host app
- [Gateway integrations for external apps](https://docs.openclaw.ai/gateway/external-apps): Current integration path for external apps, scripts, dashboards, CI jobs, and IDE extensions
- [Gateway lock](https://docs.openclaw.ai/gateway/gateway-lock): Gateway singleton guard: file lock plus WebSocket/HTTP bind
- [Health checks](https://docs.openclaw.ai/gateway/health): Health check commands and gateway health monitoring
- [Heartbeat](https://docs.openclaw.ai/gateway/heartbeat): Heartbeat polling messages and notification rules
- [Local model services](https://docs.openclaw.ai/gateway/local-model-services): Start local model servers on demand before OpenClaw model and embedding requests
- [Local models](https://docs.openclaw.ai/gateway/local-models): Run OpenClaw with hardware-aware local model setup or an existing model server
- [Gateway logging](https://docs.openclaw.ai/gateway/logging): Logging surfaces, file logs, WS log styles, and console formatting
- [Multiple gateways](https://docs.openclaw.ai/gateway/multiple-gateways): Run multiple OpenClaw Gateways on one host (isolation, ports, and profiles)
- [OpenAI chat completions](https://docs.openclaw.ai/gateway/openai-http-api): Expose an OpenAI-compatible /v1/chat/completions HTTP endpoint from the Gateway
- [OpenResponses API](https://docs.openclaw.ai/gateway/openresponses-http-api): Expose an OpenResponses-compatible /v1/responses HTTP endpoint from the Gateway
- [OpenShell](https://docs.openclaw.ai/gateway/openshell): Use OpenShell as a managed sandbox backend for OpenClaw agents
- [OpenTelemetry export](https://docs.openclaw.ai/gateway/opentelemetry): Index of the OpenClaw OpenTelemetry export documentation, one page per reader job
- [OpenTelemetry configuration](https://docs.openclaw.ai/gateway/opentelemetry/configuration): Signal toggles, the diagnostics.otel field reference, OTEL_* environment variables, and sampling and flushing
- [Model calls and exported metrics](https://docs.openclaw.ai/gateway/opentelemetry/model-calls-and-metrics): What a model-call span measures, Claude Code CLI turn fidelity, and the full exported metric catalog
- [Privacy and trace context](https://docs.openclaw.ai/gateway/opentelemetry/privacy-and-trace-context): Continue an upstream W3C trace through the Gateway WebSocket, and control what content leaves the process
- [Set up OpenTelemetry export](https://docs.openclaw.ai/gateway/opentelemetry/setup): Install and enable diagnostics-otel, see which processes export, check exporter health, and turn export off
- [Exported spans and diagnostic events](https://docs.openclaw.ai/gateway/opentelemetry/spans-and-events): The exported span catalog with its attributes, and the diagnostic event catalog behind the metrics and spans
- [Operator scopes](https://docs.openclaw.ai/gateway/operator-scopes): Operator roles, scopes, and approval-time checks for Gateway clients
- [Node pairing](https://docs.openclaw.ai/gateway/pairing): Node capability approvals: how nodes gain command exposure after device pairing
- [Session permission modes](https://docs.openclaw.ai/gateway/permission-modes): Session permission modes, workspace boundaries, and escalation reviewers
- [Portals](https://docs.openclaw.ai/gateway/portals): Expose agent-run development servers to the operator through the Gateway
- [Prometheus metrics](https://docs.openclaw.ai/gateway/prometheus): Expose OpenClaw diagnostics as Prometheus text metrics through the diagnostics-prometheus plugin
- [Gateway protocol](https://docs.openclaw.ai/gateway/protocol): Gateway WebSocket protocol: handshake, frames, versioning
- [Gateway protocol auth](https://docs.openclaw.ai/gateway/protocol/auth): Handshake auth paths, device identity, pairing signatures, and TLS pinning
- [Gateway protocol handshake](https://docs.openclaw.ai/gateway/protocol/handshake): Connect frame, hello-ok payload, client capabilities, roles, and scopes
- [Gateway protocol ledger RPCs](https://docs.openclaw.ai/gateway/protocol/ledgers): Audit ledger RPCs, scopes, cursors, and payloads
- [Gateway protocol operator methods](https://docs.openclaw.ai/gateway/protocol/operator-methods): Operator helper methods, exec approvals, and agent delivery fallback
- [Gateway protocol presence and events](https://docs.openclaw.ai/gateway/protocol/presence): Presence snapshots, node host stats, and broadcast event scoping
- [Gateway protocol session bootstrap and events](https://docs.openclaw.ai/gateway/protocol/rpc-bootstrap-and-events): Session list bootstrap, the common event families, and the node helper and exec lifecycle contracts
- [Gateway protocol device, node, and approval methods](https://docs.openclaw.ai/gateway/protocol/rpc-devices-nodes-and-approvals): Gateway RPC families for device pairing, node invoke, approvals, Control UI commands, and automation
- [Gateway protocol RPC methods](https://docs.openclaw.ai/gateway/protocol/rpc-methods): RPC method families, discovery, session list bootstrap, and event families
- [Gateway protocol session control](https://docs.openclaw.ai/gateway/protocol/rpc-session-control): The session control RPC family: listing, sending, streaming, forking, and lifecycle methods
- [Gateway protocol system and channel methods](https://docs.openclaw.ai/gateway/protocol/rpc-system-and-channels): Gateway RPC families for system status, memory, models, channels, plugins, messaging, and the operator terminal
- [Gateway protocol talk, config, and agent methods](https://docs.openclaw.ai/gateway/protocol/rpc-talk-config-and-agents): Gateway RPC families for Talk and TTS, secrets and config, updates, the wizard, and agent workspaces
- [Gateway protocol transport](https://docs.openclaw.ai/gateway/protocol/transport): Gateway WS transport: packages, frame shapes, limits, and WebRTC Talk control
- [Gateway protocol versioning](https://docs.openclaw.ai/gateway/protocol/versioning): Protocol version constants, the N-1 node window, and client defaults
- [Remote access](https://docs.openclaw.ai/gateway/remote): Remote access using Gateway WS, SSH tunnels, and tailnets
- [Restart recovery](https://docs.openclaw.ai/gateway/restart-recovery): What survives a gateway restart or crash: eligible interrupted turns resume, native subagents settle, and queued deliveries drain
- [Sandbox vs tool policy vs elevated](https://docs.openclaw.ai/gateway/sandbox-vs-tool-policy-vs-elevated): Why a tool is blocked: sandbox runtime, tool allow/deny policy, and elevated exec gates
- [Sandboxing](https://docs.openclaw.ai/gateway/sandboxing): How OpenClaw sandboxing works: modes, scopes, workspace access, and images
- [Crabbox backend](https://docs.openclaw.ai/gateway/sandboxing/crabbox-backend): Tool-call isolation on a Crabbox-leased machine with fixed runtime identity and repository-owned execution
- [Docker backend](https://docs.openclaw.ai/gateway/sandboxing/docker-backend): Docker backend defaults, the restricted runtime posture, DooD constraints, and the sandboxed browser
- [Images and setup](https://docs.openclaw.ai/gateway/sandboxing/images-and-setup): Building the default, common, and browser sandbox images, and the network and certificate defaults
- [Modes, scope, and backend](https://docs.openclaw.ai/gateway/sandboxing/modes-scope-and-backend): The three settings that control when sandboxing applies and how many environments are created
- [Multiple folders for one agent](https://docs.openclaw.ai/gateway/sandboxing/multiple-folders-for-one-agent): Docker bind mounts for extra host folders, their access modes, and the bind security rules
- [OpenShell backend](https://docs.openclaw.ai/gateway/sandboxing/openshell-backend): The OpenShell managed sandbox, its mirror and remote workspace modes, and its lifecycle
- [Podman backend](https://docs.openclaw.ai/gateway/sandboxing/podman-backend): Selecting the native Podman CLI, rootless user mapping, and Podman-outside-of-Podman constraints
- [setupCommand (one-time container setup)](https://docs.openclaw.ai/gateway/sandboxing/setup-command): The one-time container setup hook, where it is configured, and its common pitfalls
- [SSH backend](https://docs.openclaw.ai/gateway/sandboxing/ssh-backend): Sandboxing tools on an arbitrary SSH-accessible machine, and its remote-canonical workspace
- [Supported capability matrix](https://docs.openclaw.ai/gateway/sandboxing/supported-capability-matrix): What each sandbox backend supports for shell, files, workspace, network, browser, and plugin tools
- [What gets sandboxed](https://docs.openclaw.ai/gateway/sandboxing/what-gets-sandboxed): Which tool calls move into the sandbox and which stay on the Gateway host
- [Workspace access](https://docs.openclaw.ai/gateway/sandboxing/workspace-access): The none, ro, and rw workspace access modes, the role-required cap, and skill mirroring
- [Secrets management](https://docs.openclaw.ai/gateway/secrets): Secrets management: SecretRef contract, shared secret store, runtime snapshots, and safe one-way scrubbing
- [Secrets apply plan contract](https://docs.openclaw.ai/gateway/secrets-plan-contract): Contract for `secrets apply` plans: target validation, path matching, and SQLite auth-profile target scope
- [Secrets integration examples](https://docs.openclaw.ai/gateway/secrets/integration-examples): Exec provider recipes for external secret managers, MCP server environment variables, and sandbox SSH
- [Secrets operations and behavior](https://docs.openclaw.ai/gateway/secrets/operations): Supported surfaces, precedence, activation triggers, degraded signals, and the audit and configure workflow
- [Secrets runtime model](https://docs.openclaw.ai/gateway/secrets/runtime-model): How the Gateway resolves, injects, and isolates secrets at runtime
- [Shared secret store and egress proxy](https://docs.openclaw.ai/gateway/secrets/secret-store-and-egress): The shared secret store, the default-off secret egress proxy, and file-backed API keys
- [SecretRef contract and provider config](https://docs.openclaw.ai/gateway/secrets/secretref-contract): SecretRef sources, id grammar, validation rules, and provider configuration
- [Security](https://docs.openclaw.ai/gateway/security): Trust model, safe defaults, and hardening guidance for running OpenClaw
- [Access control and allowlists](https://docs.openclaw.ai/gateway/security/access-control): Who can reach the agent: DM policy, allowlists, DM session isolation, context visibility, and command authorization
- [Security audit checks](https://docs.openclaw.ai/gateway/security/audit-checks): Reference catalog of checkIds emitted by openclaw security audit
- [Browser control risks](https://docs.openclaw.ai/gateway/security/browser-control): What giving the model a real browser exposes, and the SSRF policy that bounds it
- [Dependency locking](https://docs.openclaw.ai/gateway/security/dependency-locking): How OpenClaw reviews dependency changes and packages plugin runtime dependencies
- [Gateway exposure runbook](https://docs.openclaw.ai/gateway/security/exposure-runbook): Pre-flight and rollback checklist before exposing an OpenClaw Gateway beyond loopback
- [Hardened baselines](https://docs.openclaw.ai/gateway/security/hardened-baseline): Copy/paste Gateway configurations that keep the deployment private, paired, and tool-restricted
- [Network exposure](https://docs.openclaw.ai/gateway/security/network-exposure): Bind, firewall, discovery, Gateway auth, Tailscale, reverse proxy, and Control UI exposure settings
- [Operator incident response](https://docs.openclaw.ai/gateway/security/operator-incident-response): Contain, rotate, audit, and collect evidence after a suspected compromise of your own Gateway
- [Prompt injection](https://docs.openclaw.ai/gateway/security/prompt-injection): How untrusted content reaches the model, what OpenClaw does about it, and which layers you still have to enforce
- [Rate limiting](https://docs.openclaw.ai/gateway/security/rate-limiting): Reference for every Gateway rate limit: pre-auth socket budgets and lockouts, browser and webhook throttles, the control-plane write backstop, ACP session caps, and restart cooldown
- [Running the security audit](https://docs.openclaw.ai/gateway/security/running-the-audit): What `openclaw security audit` inspects and the order to fix its findings in
- [Secrets, storage, and logs](https://docs.openclaw.ai/gateway/security/secrets-and-storage): What OpenClaw writes to disk, which files hold credentials, and how logs and transcripts are handled
- [Secure file operations](https://docs.openclaw.ai/gateway/security/secure-file-operations): How OpenClaw handles local file access safely, including native Windows credential checks
- [Tool and agent permissions](https://docs.openclaw.ai/gateway/security/tool-permissions): Cross-provider messaging, control-plane tools, node execution, plugins, sandboxing, and per-agent access profiles
- [Security trust model](https://docs.openclaw.ai/gateway/security/trust-model): The trust boundary OpenClaw supports, the controls it does and does not provide, and how to report a vulnerability
- [Give your Gateway a stable HTTPS URL](https://docs.openclaw.ai/gateway/stable-https-url): Give a loopback-only Gateway a stable, tailnet-only HTTPS URL with Tailscale Serve
- [Tailscale](https://docs.openclaw.ai/gateway/tailscale): Integrated Tailscale Serve/Funnel for the Gateway dashboard
- [Deploy a team server](https://docs.openclaw.ai/gateway/team-server): Deploy a shared team Gateway with Cloudflare Access, verified GitHub identities, roles, session sharing, and recoverable operations
- [Usage telemetry and update checks](https://docs.openclaw.ai/gateway/telemetry): Daily update requests, approximate location, optional anonymous feature statistics, and privacy controls
- [Tools invoke API](https://docs.openclaw.ai/gateway/tools-invoke-http-api): Invoke a single tool directly via the Gateway HTTP endpoint
- [Troubleshooting](https://docs.openclaw.ai/gateway/troubleshooting): Index of the deep gateway troubleshooting runbook, grouped by symptom area
- [Agent replies and Control UI](https://docs.openclaw.ai/gateway/troubleshooting/agent-replies-and-control-ui): Storage errors, missing replies, and dashboard or Control UI connectivity and auth codes
- [Channel delivery and tools](https://docs.openclaw.ai/gateway/troubleshooting/channel-delivery-and-tools): Channels that connect but do not deliver, cron and heartbeat delivery, and node or browser tool failures
- [Config validation and checks](https://docs.openclaw.ai/gateway/troubleshooting/config-validation-and-probes): Recovering from a rejected config and reading gateway check warnings
- [Gateway service and process](https://docs.openclaw.ai/gateway/troubleshooting/gateway-service-and-process): The managed gateway service not running, macOS launchd faults, and high-memory exits
- [Skills and model providers](https://docs.openclaw.ai/gateway/troubleshooting/skills-and-model-providers): Skill root containment, Anthropic 429s, upstream 403s, and local OpenAI-compatible backends
- [Updates and rollbacks](https://docs.openclaw.ai/gateway/troubleshooting/updates-and-rollbacks): Symptoms that appear after an update, a rollback, or a split-brain install of the Gateway
- [Trusted proxy auth](https://docs.openclaw.ai/gateway/trusted-proxy-auth): Delegate gateway authentication to a trusted reverse proxy (Pomerium, Caddy, nginx + OAuth)
- [Help](https://docs.openclaw.ai/help): Help hub: common fixes, install sanity, and where to look when something breaks
- [Debugging](https://docs.openclaw.ai/help/debugging): Debugging tools: watch mode, raw model streams, and tracing reasoning leakage
- [Environment variables](https://docs.openclaw.ai/help/environment): Where OpenClaw loads environment variables and the precedence order
- [FAQ](https://docs.openclaw.ai/help/faq): Frequently asked questions about OpenClaw setup, configuration, and usage
- [FAQ: first-run setup](https://docs.openclaw.ai/help/faq-first-run): FAQ: quick-start and first-run setup — install, onboard, auth, subscriptions, initial failures
- [FAQ: providers, hardware, and hosting](https://docs.openclaw.ai/help/faq-first-run/providers-and-hosting): FAQ: provider auth and limits, model choice, hardware, and where to run the Gateway
- [FAQ: quick start and first-run setup](https://docs.openclaw.ai/help/faq-first-run/quick-start): FAQ: install, onboarding, first-run failures, builds, and subscription basics
- [FAQ: models and auth](https://docs.openclaw.ai/help/faq-models): FAQ: model defaults, selection, aliases, switching, failover, and auth profiles
- [Chat commands, aborting tasks, and stopping a run](https://docs.openclaw.ai/help/faq/chat-commands-and-stopping): Hiding system messages, cancelling runs, cross-context messaging, and the steering queue
- [Config basics](https://docs.openclaw.ai/help/faq/config-basics): Config format and location, restarts, web search, config.apply recovery, and browser control
- [Env vars and .env loading](https://docs.openclaw.ai/help/faq/env-vars): How OpenClaw loads environment variables and why service starts lose them
- [Gateway ports, already running, and remote mode](https://docs.openclaw.ai/help/faq/gateway-ports-and-remote-mode): Gateway ports, remote-mode clients, bind addresses, and running more than one Gateway
- [Logging and debugging](https://docs.openclaw.ai/help/faq/logging-and-debugging): Where logs are, service start and restart, and what to check when replies never arrive
- [Media and attachments](https://docs.openclaw.ai/help/faq/media-and-attachments): Why a generated image or PDF was not delivered to chat
- [Remote gateways and nodes](https://docs.openclaw.ai/help/faq/remote-gateways-and-nodes): Command propagation, Tailscale setups, multi-device nodes, and applying config remotely
- [Sandboxing and memory](https://docs.openclaw.ai/help/faq/sandboxing-and-memory): Sandbox modes, host folder binds, and how agent memory persists
- [Security and access control](https://docs.openclaw.ai/help/faq/security-and-access-control): Exposure, prompt injection, third-party skills, pairing, and how much autonomy to grant
- [Sessions and multiple chats](https://docs.openclaw.ai/help/faq/sessions-and-chats): Fresh sessions, context truncation, full resets, groups, and multi-agent setups
- [Skills and automation](https://docs.openclaw.ai/help/faq/skills-and-automation): Customizing skills, per-task models, subagents, cron jobs, and background runs
- [What is OpenClaw?](https://docs.openclaw.ai/help/faq/what-is-openclaw): What OpenClaw is, who it is for, how it is funded, and how it compares
- [Where things live on disk](https://docs.openclaw.ai/help/faq/where-things-live-on-disk): Data locations, AGENTS.md and SOUL.md placement, backups, and uninstalling
- [Scripts](https://docs.openclaw.ai/help/scripts): Repository scripts: purpose, scope, and safety notes
- [Testing](https://docs.openclaw.ai/help/testing): Index of the OpenClaw testing kit, one page per reader job
- [Testing: live suites](https://docs.openclaw.ai/help/testing-live): Live (network-touching) tests: model matrix, CLI backends, ACP, media providers, credentials
- [ACP bind and Codex app-server lanes](https://docs.openclaw.ai/help/testing-live/acp-and-codex): The live ACP conversation-bind smoke and the plugin-owned Codex app-server harness smoke
- [CLI backend and APNs lanes](https://docs.openclaw.ai/help/testing-live/cli-backends): Live CLI backend smoke (Claude, Gemini, other local CLIs) and the APNs HTTP/2 proxy reachability check
- [OpenAI long context and the live model matrix](https://docs.openclaw.ai/help/testing-live/long-context-and-matrix): The OpenAI long-context proof runs, the recommended live recipes, and the curated model matrix
- [Media provider live lanes](https://docs.openclaw.ai/help/testing-live/media-providers): Deepgram, BytePlus, ComfyUI, and the shared image/music/video generation live sweeps
- [Live model smoke (profile keys)](https://docs.openclaw.ai/help/testing-live/model-smoke): The two-layer live model smoke: direct completion and the gateway + dev agent pipeline
- [Quick live smokes and the Android node sweep](https://docs.openclaw.ai/help/testing-live/quick-smokes): Ad hoc local media/voice smokes plus the Android node capability sweep
- [Testing: updates and plugins](https://docs.openclaw.ai/help/testing-updates-plugins): How OpenClaw validates update paths, package migrations, and plugin install/update behavior
- [Contract tests](https://docs.openclaw.ai/help/testing/contracts): Plugin and channel contract test commands, categories, and when to run them
- [Control UI protocol load](https://docs.openclaw.ai/help/testing/control-ui-load): Measure fixed-duration Control UI protocol load against an isolated Gateway
- [Docker test runners](https://docs.openclaw.ai/help/testing/docker): The Docker works-in-Linux runners, their scheduler, lanes, and env vars
- [Live and Docker/Parallels workflows](https://docs.openclaw.ai/help/testing/live-workflows): Live provider debugging lanes plus the Docker and Parallels smokes that support them
- [QA-specific runners](https://docs.openclaw.ai/help/testing/qa-runners): The qa-lab command surface, shared Convex credentials, and adding a channel to QA
- [Test suites and commands](https://docs.openclaw.ai/help/testing/suites): The unit, e2e, and live suites, which one to run, and the offline regression checks
- [Writing and adding tests](https://docs.openclaw.ai/help/testing/writing-tests): Temp-directory rules, agent reliability eval gaps, and how to add a regression
- [General troubleshooting](https://docs.openclaw.ai/help/troubleshooting): Symptom first troubleshooting hub for OpenClaw
- [OpenClaw](https://docs.openclaw.ai/): OpenClaw is an open-source AI assistant that runs on your own hardware and meets you in every chat app you already use.
- [Install](https://docs.openclaw.ai/install): Install OpenClaw - desktop app downloads, installer script, npm/pnpm/bun, from source, Docker, and more
- [Ansible](https://docs.openclaw.ai/install/ansible): Automated, hardened OpenClaw installation with Ansible, Tailscale VPN, and firewall isolation
- [Azure](https://docs.openclaw.ai/install/azure): Run OpenClaw Gateway 24/7 on an Azure Linux VM with durable state
- [Backups](https://docs.openclaw.ai/install/backups): Back up OpenClaw state: archives, per-database snapshots, scheduling, offsite copies, and continuous replication
- [Bun](https://docs.openclaw.ai/install/bun): Bun workflow for installs, package scripts, and opt-in runtime use
- [Bun compatibility](https://docs.openclaw.ai/install/bun-compatibility): Bun runtime requirements, macOS SQLite selection, limitations, and release history
- [Cloudflare Containers](https://docs.openclaw.ai/install/cloudflare): Experimental Cloudflare Worker and Container deployment with Litestream backups to R2
- [Daytona](https://docs.openclaw.ai/install/daytona): Run OpenClaw in a Daytona cloud sandbox with SSH access and signed preview URLs
- [Release channels](https://docs.openclaw.ai/install/development-channels): Stable, extended-stable, beta, and dev channels: semantics, switching, pinning, and tagging
- [DigitalOcean](https://docs.openclaw.ai/install/digitalocean): Host OpenClaw on a DigitalOcean Droplet
- [Docker](https://docs.openclaw.ai/install/docker): Optional Docker-based setup and onboarding for OpenClaw
- [Docker VM runtime](https://docs.openclaw.ai/install/docker-vm-runtime): Shared Docker VM runtime steps for long-lived OpenClaw Gateway hosts
- [Docker Compose operations and image maintenance](https://docs.openclaw.ai/install/docker/compose-operations): Day-to-day Compose commands, operational accordions, and image maintenance
- [Docker environment variables](https://docs.openclaw.ai/install/docker/environment-variables): Every environment variable the Docker setup script and Compose file accept
- [Docker networking, providers, and storage](https://docs.openclaw.ai/install/docker/networking-and-storage): Bind modes, reaching host providers, the Claude CLI backend, Bonjour, and what persists
- [Docker agent sandbox and troubleshooting](https://docs.openclaw.ai/install/docker/sandbox-and-troubleshooting): Running agent tools in Docker sandboxes, and fixing a gateway container that misbehaves
- [exe.dev](https://docs.openclaw.ai/install/exe-dev): Run OpenClaw Gateway on exe.dev (VM + HTTPS proxy) for remote access
- [Fly.io](https://docs.openclaw.ai/install/fly): Step-by-step Fly.io deployment for OpenClaw with persistent storage and HTTPS
- [GCP](https://docs.openclaw.ai/install/gcp): Run OpenClaw Gateway 24/7 on a GCP Compute Engine VM with Docker
- [Hetzner](https://docs.openclaw.ai/install/hetzner): Run OpenClaw Gateway 24/7 on a Hetzner VPS with Docker
- [Hostinger](https://docs.openclaw.ai/install/hostinger): Host OpenClaw on Hostinger
- [Installer internals](https://docs.openclaw.ai/install/installer): How the installer scripts work (install.sh, install-cli.sh, install.ps1), flags, and automation
- [Kubernetes](https://docs.openclaw.ai/install/kubernetes): Deploy OpenClaw Gateway to a Kubernetes cluster with Kustomize
- [macOS VMs](https://docs.openclaw.ai/install/macos-vm): Run OpenClaw in a sandboxed macOS VM (local or hosted) when you need isolation or iMessage
- [Migration guide](https://docs.openclaw.ai/install/migrating): Migration hub: cross-system imports, machine-to-machine moves, and plugin upgrades
- [Migrating from Claude](https://docs.openclaw.ai/install/migrating-claude): Move Claude Code and Claude Desktop local state into OpenClaw with a previewed import
- [Migrating from Hermes](https://docs.openclaw.ai/install/migrating-hermes): Move from Hermes to OpenClaw with a previewed, reversible import
- [Nix](https://docs.openclaw.ai/install/nix): Install OpenClaw declaratively with Nix
- [Node.js](https://docs.openclaw.ai/install/node): Install and configure Node.js for OpenClaw - version requirements, install options, and PATH troubleshooting
- [Node.js compatibility](https://docs.openclaw.ai/install/node-compatibility): Supported Node.js versions, SQLite requirements, platform limits, and release history
- [Northflank](https://docs.openclaw.ai/install/northflank): Deploy OpenClaw on Northflank with one-click template
- [Oracle Cloud](https://docs.openclaw.ai/install/oracle): Host OpenClaw on Oracle Cloud's Always Free ARM tier
- [Podman](https://docs.openclaw.ai/install/podman): Run OpenClaw in a rootless Podman container
- [Railway](https://docs.openclaw.ai/install/railway): Deploy OpenClaw on Railway with one-click template
- [Raspberry Pi](https://docs.openclaw.ai/install/raspberry-pi): Host OpenClaw on a Raspberry Pi for always-on self-hosting
- [Render](https://docs.openclaw.ai/install/render): Deploy OpenClaw on Render with Infrastructure-as-Code
- [Uninstall](https://docs.openclaw.ai/install/uninstall): Uninstall OpenClaw completely (CLI, service, state, workspace)
- [Update troubleshooting](https://docs.openclaw.ai/install/update-troubleshooting): Recover from failed OpenClaw updates in the Control UI or CLI
- [Updating](https://docs.openclaw.ai/install/updating): Updating OpenClaw safely (global install or source), plus rollback strategy
- [Automatic updates](https://docs.openclaw.ai/install/updating/automatic-updates): The auto-updater, per-channel automatic behavior, and how update campaigns apply and report an update
- [Rollback and recovery](https://docs.openclaw.ai/install/updating/rollback-and-recovery): Downgrading, automatic schema-neutral rollback, verified pre-update backups, and triage when an update leaves you stuck
- [Other update methods](https://docs.openclaw.ai/install/updating/update-methods): Switching install types, the source-server script, re-running the installer, and manual npm/pnpm/bun updates
- [Upstash Box](https://docs.openclaw.ai/install/upstash): Host OpenClaw on Upstash Box with keep-alive and SSH tunnel access
- [Logging](https://docs.openclaw.ai/logging): File logs, console output, CLI tailing, and the Control UI Logs tab
- [Maturity scorecard](https://docs.openclaw.ai/maturity/scorecard): OpenClaw release readiness scores for product areas, integrations, and supported workflows.
- [Maturity taxonomy](https://docs.openclaw.ai/maturity/taxonomy): Detailed reference for the product areas and checks behind the OpenClaw maturity scorecard.
- [Network](https://docs.openclaw.ai/network): Network hub: gateway surfaces, pairing, discovery, and security
- [Nodes](https://docs.openclaw.ai/nodes): Nodes: pairing, capabilities, permissions, and CLI helpers for camera/screen/device/notifications/system and the macOS widget panel
- [Audio and voice notes](https://docs.openclaw.ai/nodes/audio): How inbound audio/voice notes are downloaded, transcribed, and injected into replies
- [Camera capture](https://docs.openclaw.ai/nodes/camera): Camera capture and macOS physical PTZ control on paired nodes
- [Node command policy](https://docs.openclaw.ai/nodes/command-policy): Platform default allowlists, dangerous-command opt-ins, and the gateway.nodes configuration
- [Computer use](https://docs.openclaw.ai/nodes/computer-use): Capability-based control of Gateway and paired node desktops through the computer tool
- [Node device commands](https://docs.openclaw.ai/nodes/device-commands): CLI helpers for the widget panel, camera, screen, location, SMS, and device data
- [Node file transfers](https://docs.openclaw.ai/nodes/file-transfers): Upload files into a node terminal and move files with the File Transfer plugin
- [Image and media support](https://docs.openclaw.ai/nodes/images): Image and media handling rules for send, gateway, and agent replies
- [Location command](https://docs.openclaw.ai/nodes/location-command): Location command for nodes, platform permission modes, and Linux GeoClue setup
- [Node-hosted MCP servers and skills](https://docs.openclaw.ai/nodes/mcp-and-skills): Publish MCP servers, skills, and local model inference from a paired node
- [Media playback](https://docs.openclaw.ai/nodes/media-playback): Inline audio and video playback across the Control UI and native apps
- [Media understanding](https://docs.openclaw.ai/nodes/media-understanding): Inbound image/audio/video understanding (optional) with provider + CLI fallbacks
- [Run commands on a node](https://docs.openclaw.ai/nodes/node-exec): Allowlist node commands, point exec at a node, and invoke commands directly
- [Run a node host](https://docs.openclaw.ai/nodes/node-host): Start a headless or service node host, pair it, and use its system commands
- [Node pairing and status](https://docs.openclaw.ai/nodes/pairing-and-status): Pair a node to the Gateway, read its status, and upgrade a fleet in the right order
- [Active computer presence](https://docs.openclaw.ai/nodes/presence): Detect the Mac you most recently used and route node alerts there
- [Node session catalogs](https://docs.openclaw.ai/nodes/session-catalogs): Discover, import, and continue native session transcripts on the Gateway and paired nodes
- [Host OpenClaw sessions on a node](https://docs.openclaw.ai/nodes/session-hosting): Enable worker session hosting on a paired node, choose a device, and isolate workers in containers
- [Talk mode](https://docs.openclaw.ai/nodes/talk): Talk mode: continuous speech conversations across local STT/TTS and realtime voice
- [Talk client UI](https://docs.openclaw.ai/nodes/talk/client-ui): Talk controls and behavior in the macOS, Apple Watch, and Android clients
- [Talk on macOS and the Gateway relay](https://docs.openclaw.ai/nodes/talk/macos-relay): macOS Talk overlay behavior and the streamed realtime Gateway relay path
- [Talk realtime sessions and delegation](https://docs.openclaw.ai/nodes/talk/realtime-sessions): Voice selection from chat, realtime delegation, steering, transcripts, and browser Talk behavior
- [Talk session ownership](https://docs.openclaw.ai/nodes/talk/session-ownership): How Talk resolves the owning agent, session key, store, and control authority
- [Node troubleshooting](https://docs.openclaw.ai/nodes/troubleshooting): Troubleshoot node pairing, foreground requirements, permissions, and tool failures
- [Voice wake](https://docs.openclaw.ai/nodes/voicewake): Global voice wake words (Gateway-owned) and how they sync across nodes
- [OpenClaw agent runtime workflow](https://docs.openclaw.ai/openclaw-agent-runtime): Developer workflow for OpenClaw agent runtime: build, test, and live validation
- [Platforms](https://docs.openclaw.ai/platforms): Platform support overview (Gateway + companion apps)
- [Android app](https://docs.openclaw.ai/platforms/android): Android app (node): pairing, connection recovery, chat, voice, and device commands
- [ChromeOS](https://docs.openclaw.ai/platforms/chromeos): Run the OpenClaw Gateway on ChromeOS inside a Crostini Linux container
- [EasyRunner](https://docs.openclaw.ai/platforms/easyrunner): Run the OpenClaw Gateway on EasyRunner with Podman and Caddy
- [iOS app](https://docs.openclaw.ai/platforms/ios): iOS node app: connect to the Gateway, pairing, device capabilities, and troubleshooting
- [HealthKit summaries](https://docs.openclaw.ai/platforms/ios-healthkit): Enable and invoke privacy-gated HealthKit summaries from an iOS node
- [Linux app](https://docs.openclaw.ai/platforms/linux): Linux support + companion app status
- [Gateway on macOS](https://docs.openclaw.ai/platforms/mac/bundled-gateway): Bundled Gateway runtime, app hosting, and background services on macOS
- [Widget panel](https://docs.openclaw.ai/platforms/mac/canvas): Present hosted widgets in the macOS panel
- [macOS dev setup](https://docs.openclaw.ai/platforms/mac/dev-setup): Setup guide for developers working on the OpenClaw macOS app
- [Health checks (macOS)](https://docs.openclaw.ai/platforms/mac/health): How the macOS app reports gateway/channel health states
- [Menu bar icon](https://docs.openclaw.ai/platforms/mac/icon): Menu bar icon states and animations for OpenClaw on macOS
- [macOS logging](https://docs.openclaw.ai/platforms/mac/logging): macOS app log redaction, opt-in diagnostics files, and native OSLog privacy flags
- [Menu bar](https://docs.openclaw.ai/platforms/mac/menu-bar): Menu bar status logic and what is surfaced to users
- [Peekaboo bridge](https://docs.openclaw.ai/platforms/mac/peekaboo): PeekabooBridge integration for macOS UI automation
- [macOS permissions](https://docs.openclaw.ai/platforms/mac/permissions): macOS permission persistence (TCC) and signing requirements
- [Remote control](https://docs.openclaw.ai/platforms/mac/remote): macOS app flow for controlling a remote OpenClaw Gateway
- [macOS signing](https://docs.openclaw.ai/platforms/mac/signing): Signing steps for macOS debug builds generated by packaging scripts
- [Skills (macOS)](https://docs.openclaw.ai/platforms/mac/skills): Manage Gateway skills from the macOS app's Dashboard
- [Voice overlay](https://docs.openclaw.ai/platforms/mac/voice-overlay): Voice overlay lifecycle when wake-word and push-to-talk overlap
- [Voice wake (macOS)](https://docs.openclaw.ai/platforms/mac/voicewake): Voice wake and push-to-talk modes plus routing details in the mac app
- [WebChat (macOS)](https://docs.openclaw.ai/platforms/mac/webchat): Choose the Web or experimental Native Mac experience and use Gateway chat windows
- [macOS IPC](https://docs.openclaw.ai/platforms/mac/xpc): macOS IPC architecture for OpenClaw app, gateway node transport, and PeekabooBridge
- [macOS app](https://docs.openclaw.ai/platforms/macos): Install and use the OpenClaw macOS menu bar app
- [Omarchy](https://docs.openclaw.ai/platforms/omarchy): OpenClaw on Omarchy: bar plugin, desktop app handoff, and support
- [Windows](https://docs.openclaw.ai/platforms/windows): Windows support: Windows Hub, native CLI and Gateway, WSL2 gateway setup, node mode, and troubleshooting
- [Adding capabilities (contributor guide)](https://docs.openclaw.ai/plugins/adding-capabilities): Contributor guide for adding a new shared capability to the OpenClaw plugin system
- [Admin HTTP RPC plugin](https://docs.openclaw.ai/plugins/admin-http-rpc): Expose selected Gateway control-plane methods through the bundled, opt-in admin-http-rpc plugin
- [Agents API](https://docs.openclaw.ai/plugins/agentsapi): Use the cloud-hosted Agents API harness, powered by Codex, with OpenClaw
- [Apple Foundation Models](https://docs.openclaw.ai/plugins/apple-fm): Use on-device Apple Foundation Models for lightweight OpenClaw setup on a Mac
- [Plugin internals](https://docs.openclaw.ai/plugins/architecture): Plugin internals: capability model, ownership, contracts, load pipeline, and runtime helpers
- [Plugin architecture internals](https://docs.openclaw.ai/plugins/architecture-internals): Plugin architecture internals: load pipeline, registry, runtime hooks, HTTP routes, and reference tables
- [Channel surfaces](https://docs.openclaw.ai/plugins/architecture-internals/channel-surfaces): Conversation binding callbacks, message tool schemas, target resolution, config-backed directories, and read-only inspection
- [Context engines](https://docs.openclaw.ai/plugins/architecture-internals/context-engines): Registering a context engine plugin that owns session ingest, assembly, and compaction
- [Gateway routes](https://docs.openclaw.ai/plugins/architecture-internals/gateway-routes): Registering plugin HTTP routes on the Gateway, and their auth, scope, and lifecycle rules
- [Plugin load pipeline and registry](https://docs.openclaw.ai/plugins/architecture-internals/load-pipeline): How OpenClaw discovers, gates, loads, and registers plugins, and what the plugin cache holds
- [Adding a new capability](https://docs.openclaw.ai/plugins/architecture-internals/new-capability): The sequence, file checklist, and contract-test pattern for adding a capability to the plugin system
- [Package packs and import paths](https://docs.openclaw.ai/plugins/architecture-internals/packaging): Plugin SDK import subpaths, multi-extension package packs, and channel catalog and install metadata
- [Provider runtime hooks and catalogs](https://docs.openclaw.ai/plugins/architecture-internals/provider-hooks): The provider hook order table, worked provider example, bundled hook shapes, and model catalog registration
- [Core runtime helpers](https://docs.openclaw.ai/plugins/architecture-internals/runtime-helpers): Calling core speech, media understanding, subagent, web search, and image generation helpers from a plugin
- [Beam plugin](https://docs.openclaw.ai/plugins/beam): Publish redacted coding sessions for shared viewing and independent Team continuation
- [Building plugins](https://docs.openclaw.ai/plugins/building-plugins): Create your first OpenClaw plugin in minutes
- [Plugin bundles](https://docs.openclaw.ai/plugins/bundles): Install and use Agent Plugins, Codex, Claude, and Cursor bundles as OpenClaw plugins
- [Building CLI backend plugins](https://docs.openclaw.ai/plugins/cli-backend-plugins): Build a plugin that registers a local AI CLI backend
- [Cloudflare](https://docs.openclaw.ai/plugins/cloudflare): Configure Cloudflare R2 as an encrypted OpenClaw storage location
- [Codex Computer Use](https://docs.openclaw.ai/plugins/codex-computer-use): Set up Codex Computer Use for Codex-mode OpenClaw agents
- [Codex harness](https://docs.openclaw.ai/plugins/codex-harness): Run OpenClaw embedded agent turns through the official Codex app-server harness
- [Codex harness reference](https://docs.openclaw.ai/plugins/codex-harness-reference): Configuration, auth, discovery, and app-server reference for the Codex harness
- [Codex app-server transport](https://docs.openclaw.ai/plugins/codex-harness-reference/app-server-transport): App-server transport selection, the appServer field table, and local testing env overrides
- [Codex approval and sandbox modes](https://docs.openclaw.ai/plugins/codex-harness-reference/approval-and-sandbox): YOLO and guardian approval presets, and sandboxed native execution paths
- [Codex auth and environment isolation](https://docs.openclaw.ai/plugins/codex-harness-reference/auth): Codex auth selection order, credential handling, and environment isolation
- [Codex dynamic tools](https://docs.openclaw.ai/plugins/codex-harness-reference/dynamic-tools): How OpenClaw dynamic tools are exposed to Codex app-server turns
- [Codex model discovery](https://docs.openclaw.ai/plugins/codex-harness-reference/model-discovery): Codex app-server model discovery, offline hints, and catalog rules
- [Codex restricted turns](https://docs.openclaw.ai/plugins/codex-harness-reference/restricted-turns): When a tool policy restricts the Codex native surface, and what ring zero adds
- [Codex session catalog and supervision](https://docs.openclaw.ai/plugins/codex-harness-reference/supervision): Native Codex session discovery and agent-facing supervision settings
- [Codex timeouts and turn settlement](https://docs.openclaw.ai/plugins/codex-harness-reference/timeouts): Dynamic tool timeout order, turn execution budgets, and local settlement
- [Codex workspace bootstrap files](https://docs.openclaw.ai/plugins/codex-harness-reference/workspace-bootstrap-files): Which workspace bootstrap files reach a Codex turn, and how they are carried
- [Codex harness runtime](https://docs.openclaw.ai/plugins/codex-harness-runtime): Runtime boundaries, hooks, tools, permissions, and diagnostics for the Codex harness
- [Codex compaction and transcript mirror](https://docs.openclaw.ai/plugins/codex-harness-runtime/compaction): Native Codex compaction, the OpenClaw transcript mirror, and continuity projection
- [Codex hook boundaries](https://docs.openclaw.ai/plugins/codex-harness-runtime/hooks): Which hook layer owns each Codex turn event, and what the native hook relay can do
- [Codex native permissions and elicitations](https://docs.openclaw.ai/plugins/codex-harness-runtime/permissions): Native permission decisions, remembered approvals, and MCP elicitation limits
- [Codex queue steering and feedback upload](https://docs.openclaw.ai/plugins/codex-harness-runtime/queue-and-feedback): Active-run queue steering on Codex turns and the Codex feedback upload path
- [Codex process recovery](https://docs.openclaw.ai/plugins/codex-harness-runtime/recovery): Orphaned Codex app-server process detection and cleanup after a hard Gateway stop
- [Codex replies and final answers](https://docs.openclaw.ai/plugins/codex-harness-runtime/replies): Visible reply delivery, heartbeat turns, and bounded final-answer recovery
- [Codex sandbox process streaming](https://docs.openclaw.ai/plugins/codex-harness-runtime/sandbox-streaming): Experimental native sandbox execution streaming and node-backed remote exec
- [Codex thread bindings and supervision](https://docs.openclaw.ai/plugins/codex-harness-runtime/threads): How OpenClaw binds native Codex threads, changes models, and continues supervised sessions
- [Codex runtime v1 support contract](https://docs.openclaw.ai/plugins/codex-harness-runtime/v1-support-contract): What is and is not supported in Codex runtime v1, with the reason for each boundary
- [Codex app-server policy](https://docs.openclaw.ai/plugins/codex-harness/app-server): App-server transport, approval posture, auth order, and environment isolation
- [Codex commands and diagnostics](https://docs.openclaw.ai/plugins/codex-harness/commands): The /codex command surface, Fast mode controls, and local thread inspection
- [Codex plugin config fields](https://docs.openclaw.ai/plugins/codex-harness/config-fields): Top-level and appServer config fields for the Codex plugin
- [Codex harness configuration](https://docs.openclaw.ai/plugins/codex-harness/configuration): Codex harness config map, restricted turns, project instructions, compaction, and long context
- [Native Codex state and features](https://docs.openclaw.ai/plugins/codex-harness/native-features): Share native Codex threads, supervise sessions, and enable native plugins and Computer Use
- [Run Codex on another machine](https://docs.openclaw.ai/plugins/codex-harness/placement): Place Codex native execution on a paired device or a cloud worker
- [Codex routing and deployment](https://docs.openclaw.ai/plugins/codex-harness/routing): Choose which OpenAI routes select Codex and shape the deployment around them
- [Codex runtime behavior](https://docs.openclaw.ai/plugins/codex-harness/runtime-behavior): Dynamic tools, web search, image loading, turn liveness, and runtime boundaries
- [Codex harness troubleshooting](https://docs.openclaw.ai/plugins/codex-harness/troubleshooting): Symptoms and fixes for Codex harness selection, app-server, and memory problems
- [Native Codex plugins](https://docs.openclaw.ai/plugins/codex-native-plugins): Configure native Codex plugins for Codex-mode OpenClaw agents
- [Supervise Codex sessions](https://docs.openclaw.ai/plugins/codex-supervision): Browse non-archived native Codex sessions and paginated transcripts across OpenClaw nodes
- [Community plugins](https://docs.openclaw.ai/plugins/community): Find and publish community-maintained OpenClaw plugins
- [Plugin compatibility](https://docs.openclaw.ai/plugins/compatibility): Plugin compatibility contracts, deprecation metadata, and migration expectations
- [Copilot SDK harness](https://docs.openclaw.ai/plugins/copilot): Run OpenClaw embedded agent turns through the external GitHub Copilot SDK harness
- [Plugin dependency resolution](https://docs.openclaw.ai/plugins/dependency-resolution): How OpenClaw installs plugin packages and resolves plugin dependencies
- [FaceTime plugin](https://docs.openclaw.ai/plugins/facetime): Set up FaceTime voice calls with your OpenClaw agent on a Mac
- [FaceTime recovery and removal](https://docs.openclaw.ai/plugins/facetime-recovery): Remove FaceTime native artifacts and restore standard Mac security policy
- [Feature plugins](https://docs.openclaw.ai/plugins/feature-plugins): Build plugins with typed operations, native pages, and replaceable Control UI views
- [Geolocation plugin](https://docs.openclaw.ai/plugins/geolocation): Resolve a connecting client's IP address to a coarse city using a locally cached database, with no per-lookup third-party calls
- [GitHub](https://docs.openclaw.ai/plugins/github): Read public GitHub issues, pull requests, commits, and comments beside chat
- [Google Meet plugin](https://docs.openclaw.ai/plugins/google-meet): Google Meet plugin: join explicit Meet URLs through Chrome or Twilio with agent talk-back defaults
- [Google Meet configuration](https://docs.openclaw.ai/plugins/google-meet/config): Google Meet plugin config defaults, overrides, and provider examples
- [Google Meet OAuth and artifacts](https://docs.openclaw.ai/plugins/google-meet/oauth-and-artifacts): Google Meet OAuth setup, refresh tokens, preflight, and conference artifact exports
- [Google Meet tool and modes](https://docs.openclaw.ai/plugins/google-meet/tool-and-modes): The google_meet tool actions, status fields, and agent versus bidi talk-back modes
- [Google Meet transports and hosts](https://docs.openclaw.ai/plugins/google-meet/transports): Google Meet transports: local Chrome, Chrome on a paired node, and Twilio dial-in
- [Google Meet troubleshooting](https://docs.openclaw.ai/plugins/google-meet/troubleshooting): Google Meet live test checklist and fixes for join, speech, creation, and Twilio failures
- [Plugin hooks](https://docs.openclaw.ai/plugins/hooks): Plugin hooks: intercept agent, tool, message, session, and Gateway lifecycle events
- [Gateway and install lifecycle hooks](https://docs.openclaw.ai/plugins/hooks/lifecycle): Plugin install checks, Gateway start and stop, and Gateway-owned cron events
- [Message and delivery hooks](https://docs.openclaw.ai/plugins/hooks/messages): Claim inbound messages and rewrite or cancel outbound deliveries
- [Prompt and session hooks](https://docs.openclaw.ai/plugins/hooks/prompt-and-session): Override the model, shape the prompt, and persist plugin-owned session state
- [Hook reference](https://docs.openclaw.ai/plugins/hooks/reference): How plugin hooks register and execute, plus the complete typed hook catalog
- [Tool call policy hooks](https://docs.openclaw.ai/plugins/hooks/tool-policy): Gate, rewrite, and approve tool calls, and rewrite tool results before persistence
- [Plugin install overrides](https://docs.openclaw.ai/plugins/install-overrides): Test packaged plugin overrides with setup-time install flows
- [llama.cpp Provider](https://docs.openclaw.ai/plugins/llama-cpp): Run GGUF chat with managed or existing llama.cpp servers and managed local embeddings
- [Logbook plugin](https://docs.openclaw.ai/plugins/logbook): Optional automatic work journal built from periodic screen snapshots
- [Manage plugins](https://docs.openclaw.ai/plugins/manage-plugins): Manage OpenClaw plugins from the Control UI or CLI
- [Plugin manifest](https://docs.openclaw.ai/plugins/manifest): Plugin manifest + JSON schema requirements (strict config validation)
- [Manifest capability fields](https://docs.openclaw.ai/plugins/manifest/capabilities): Manifest capability ownership, tool availability metadata, and activation planning
- [Manifest config and secret fields](https://docs.openclaw.ai/plugins/manifest/config-and-secrets): Manifest dangerous-flag, SecretRef migration, and secret provider preset metadata
- [Manifest model fields](https://docs.openclaw.ai/plugins/manifest/models): Manifest model catalog, shorthand family, id normalization, and pricing fields
- [Manifest versus package.json](https://docs.openclaw.ai/plugins/manifest/package-json): Which pre-runtime metadata lives in package.json, and which duplicate plugin id wins
- [Manifest provider fields](https://docs.openclaw.ai/plugins/manifest/providers): Manifest generation, media-understanding, endpoint, and request provider metadata
- [Manifest setup and auth fields](https://docs.openclaw.ai/plugins/manifest/setup-and-auth): Manifest setup descriptors, auth choices, conversation discovery, and config UI hints
- [Manifest host surface fields](https://docs.openclaw.ai/plugins/manifest/surfaces): Manifest fields for icons, themes, CLI, MCP, Control UI, dashboard, QA, channel, and backup surfaces
- [Meeting plugins](https://docs.openclaw.ai/plugins/meeting-plugins): Choose and configure Google Meet, Microsoft Teams, Slack huddles, or Zoom participation
- [Memory LanceDB](https://docs.openclaw.ai/plugins/memory-lancedb): Configure the official external LanceDB memory plugin, including local Ollama-compatible embeddings
- [Memory wiki](https://docs.openclaw.ai/plugins/memory-wiki): memory-wiki: compiled knowledge vault with provenance, claims, dashboards, and bridge mode
- [Message presentation](https://docs.openclaw.ai/plugins/message-presentation): Semantic message cards, charts, tables, controls, fallback text, and delivery hints for channel plugins
- [OC Path plugin](https://docs.openclaw.ai/plugins/oc-path): Bundled `oc-path` plugin: ships the `openclaw path` CLI for the `oc://` workspace-file addressing scheme
- [1Password plugin](https://docs.openclaw.ai/plugins/onepassword): Resolve SecretRefs and give agents curated, audited access to 1Password
- [Local ONNX decision models](https://docs.openclaw.ai/plugins/onnx): Run typed decisions locally with verified ONNX classifiers
- [Plugin inventory](https://docs.openclaw.ai/plugins/plugin-inventory): Generated inventory of OpenClaw plugins shipped in core, published externally, or kept source-only
- [Plugin permission requests](https://docs.openclaw.ai/plugins/plugin-permission-requests): Ask users to approve plugin tool calls and plugin-owned permission prompts
- [Plugin reference](https://docs.openclaw.ai/plugins/reference): Pointer to the generated OpenClaw plugin reference pages
- [A2a plugin reference](https://docs.openclaw.ai/plugins/reference/a2a): A2A v1.0 Agent-to-Agent protocol channel plugin.
- [ACPx plugin reference](https://docs.openclaw.ai/plugins/reference/acpx): OpenClaw ACP runtime backend with plugin-owned session and transport management.
- [Active Memory plugin reference](https://docs.openclaw.ai/plugins/reference/active-memory): Runs bounded pre-reply memory retrieval and implements per-agent Remember across conversations for eligible private conversations.
- [Admin Http Rpc plugin reference](https://docs.openclaw.ai/plugins/reference/admin-http-rpc): OpenClaw admin HTTP RPC endpoint.
- [Agentsapi plugin reference](https://docs.openclaw.ai/plugins/reference/agentsapi): OpenAI Agents API harness with hosted or self-hosted sessions.
- [Alibaba plugin reference](https://docs.openclaw.ai/plugins/reference/alibaba): Adds video generation provider support.
- [Amazon Bedrock plugin reference](https://docs.openclaw.ai/plugins/reference/amazon-bedrock): OpenClaw Amazon Bedrock provider plugin with model discovery, embeddings, and guardrail support.
- [Amazon Bedrock Mantle plugin reference](https://docs.openclaw.ai/plugins/reference/amazon-bedrock-mantle): OpenClaw Amazon Bedrock Mantle provider plugin for OpenAI-compatible model routing.
- [Anthropic plugin reference](https://docs.openclaw.ai/plugins/reference/anthropic): Anthropic models, Claude CLI, and native Claude session catalog.
- [Anthropic Vertex plugin reference](https://docs.openclaw.ai/plugins/reference/anthropic-vertex): OpenClaw Anthropic Vertex provider plugin for Claude models on Google Vertex AI.
- [Apple Fm plugin reference](https://docs.openclaw.ai/plugins/reference/apple-fm): On-device Apple Intelligence inference for lightweight setup and short tasks.
- [Arcee plugin reference](https://docs.openclaw.ai/plugins/reference/arcee): Adds Arcee model provider support to OpenClaw.
- [Azure Speech plugin reference](https://docs.openclaw.ai/plugins/reference/azure-speech): Azure AI Speech text-to-speech (MP3, native Ogg/Opus voice notes, PCM telephony).
- [Baseten plugin reference](https://docs.openclaw.ai/plugins/reference/baseten): OpenClaw Baseten provider plugin.
- [Beam plugin reference](https://docs.openclaw.ai/plugins/reference/beam): Read-only coding-session Beam receiver.
- [Bonjour plugin reference](https://docs.openclaw.ai/plugins/reference/bonjour): Advertise the local OpenClaw gateway over Bonjour/mDNS.
- [Brave plugin reference](https://docs.openclaw.ai/plugins/reference/brave): OpenClaw Brave Search provider plugin for web search.
- [Browser plugin reference](https://docs.openclaw.ai/plugins/reference/browser): Adds agent-callable tools.
- [Buzz plugin reference](https://docs.openclaw.ai/plugins/reference/buzz): Connect OpenClaw agents to Buzz rooms.
- [BytePlus plugin reference](https://docs.openclaw.ai/plugins/reference/byteplus): Adds BytePlus, BytePlus Plan model provider support to OpenClaw.
- [Canvas plugin reference](https://docs.openclaw.ai/plugins/reference/canvas): Presents hosted widget documents on paired macOS panels.
- [Cerebras plugin reference](https://docs.openclaw.ai/plugins/reference/cerebras): Adds Cerebras model provider support to OpenClaw.
- [Chutes plugin reference](https://docs.openclaw.ai/plugins/reference/chutes): Adds Chutes model provider support to OpenClaw.
- [ClawRouter plugin reference](https://docs.openclaw.ai/plugins/reference/clawrouter): Adds ClawRouter model provider support to OpenClaw.
- [Clickclack plugin reference](https://docs.openclaw.ai/plugins/reference/clickclack): OpenClaw ClickClack channel plugin.
- [Cloudflare plugin reference](https://docs.openclaw.ai/plugins/reference/cloudflare): Cloudflare R2 storage for named OpenClaw storage locations.
- [Cloudflare AI Gateway plugin reference](https://docs.openclaw.ai/plugins/reference/cloudflare-ai-gateway): Adds Cloudflare AI Gateway model provider support to OpenClaw.
- [Code Mode Quickjs plugin reference](https://docs.openclaw.ai/plugins/reference/code-mode-quickjs): Hardened JavaScript execution for Code Mode using QuickJS in WebAssembly.
- [Codex plugin reference](https://docs.openclaw.ai/plugins/reference/codex): Codex app-server harness and native session catalog.
- [Cohere plugin reference](https://docs.openclaw.ai/plugins/reference/cohere): OpenClaw Cohere provider plugin.
- [ComfyUI plugin reference](https://docs.openclaw.ai/plugins/reference/comfy): Adds ComfyUI model provider support to OpenClaw.
- [Copilot plugin reference](https://docs.openclaw.ai/plugins/reference/copilot): Registers the GitHub Copilot agent runtime.
- [Copilot Proxy plugin reference](https://docs.openclaw.ai/plugins/reference/copilot-proxy): Adds Copilot Proxy model provider support to OpenClaw.
- [Crabbox plugin reference](https://docs.openclaw.ai/plugins/reference/crabbox): Cloud worker provider and lease-backed sandbox backend for the Crabbox CLI.
- [Cua Computer plugin reference](https://docs.openclaw.ai/plugins/reference/cua-computer): Experimental CUA Driver computer control for macOS, Windows, and Linux node hosts.
- [Deepgram plugin reference](https://docs.openclaw.ai/plugins/reference/deepgram): Deepgram audio transcription with Nova and Flux models, plus realtime speech recognition.
- [DeepInfra plugin reference](https://docs.openclaw.ai/plugins/reference/deepinfra): Adds DeepInfra model provider support to OpenClaw.
- [DeepSeek plugin reference](https://docs.openclaw.ai/plugins/reference/deepseek): Adds DeepSeek model provider support to OpenClaw.
- [Device Pair plugin reference](https://docs.openclaw.ai/plugins/reference/device-pair): Generate setup codes and approve device pairing requests.
- [Diagnostics OpenTelemetry plugin reference](https://docs.openclaw.ai/plugins/reference/diagnostics-otel): OpenClaw diagnostics OpenTelemetry exporter for metrics, traces, and logs.
- [Diagnostics Prometheus plugin reference](https://docs.openclaw.ai/plugins/reference/diagnostics-prometheus): OpenClaw diagnostics Prometheus exporter for runtime metrics.
- [Diffs plugin reference](https://docs.openclaw.ai/plugins/reference/diffs): OpenClaw read-only diff viewer plugin and file renderer for agents.
- [Diffs Language Pack plugin reference](https://docs.openclaw.ai/plugins/reference/diffs-language-pack): Adds syntax highlighting for languages outside the default diffs viewer set.
- [Discord plugin reference](https://docs.openclaw.ai/plugins/reference/discord): OpenClaw Discord channel plugin for channels, DMs, commands, and app events.
- [Document Extract plugin reference](https://docs.openclaw.ai/plugins/reference/document-extract): Extract text and fallback page images from local document attachments.
- [DuckDuckGo plugin reference](https://docs.openclaw.ai/plugins/reference/duckduckgo): Adds web search provider support.
- [Elevenlabs plugin reference](https://docs.openclaw.ai/plugins/reference/elevenlabs): Adds media understanding provider support. Adds realtime transcription provider support. Adds text-to-speech provider support.
- [Exa plugin reference](https://docs.openclaw.ai/plugins/reference/exa): Adds web search provider support.
- [Facetime plugin reference](https://docs.openclaw.ai/plugins/reference/facetime): Experimental FaceTime realtime voice carrier for OpenClaw agents.
- [fal plugin reference](https://docs.openclaw.ai/plugins/reference/fal): Adds fal model provider support to OpenClaw.
- [Featherless plugin reference](https://docs.openclaw.ai/plugins/reference/featherless): OpenClaw Featherless AI provider plugin.
- [Feishu plugin reference](https://docs.openclaw.ai/plugins/reference/feishu): OpenClaw Feishu/Lark channel plugin for chats and workplace tools (community maintained by @m1heng).
- [File Transfer plugin reference](https://docs.openclaw.ai/plugins/reference/file-transfer): Fetch, list, and write files on paired nodes via dedicated node commands. Bypasses bash stdout truncation by using base64 over node.invoke for binaries up to 16 MB.
- [Firecrawl plugin reference](https://docs.openclaw.ai/plugins/reference/firecrawl): Adds agent-callable tools. Adds web fetch provider support. Adds web search provider support.
- [Fireworks plugin reference](https://docs.openclaw.ai/plugins/reference/fireworks): Adds Fireworks model provider support to OpenClaw.
- [Fish Audio Speech plugin reference](https://docs.openclaw.ai/plugins/reference/fish-audio-speech): Fish Audio S2.1 hosted text-to-speech with streaming, voice notes, and telephony output.
- [Geolocation plugin reference](https://docs.openclaw.ai/plugins/reference/geolocation): Resolves client IP addresses to a coarse city using a locally cached IP-geolocation database.
- [GitHub plugin reference](https://docs.openclaw.ai/plugins/reference/github): Public GitHub link previews and document reader.
- [GitHub Copilot plugin reference](https://docs.openclaw.ai/plugins/reference/github-copilot): Adds GitHub Copilot model provider support to OpenClaw.
- [Gmi plugin reference](https://docs.openclaw.ai/plugins/reference/gmi): OpenClaw GMI Cloud provider plugin.
- [Google plugin reference](https://docs.openclaw.ai/plugins/reference/google): Adds Google, Google Gemini CLI, Google Vertex model provider support to OpenClaw.
- [Google Meet plugin reference](https://docs.openclaw.ai/plugins/reference/google-meet): OpenClaw Google Meet participant plugin for joining calls through Chrome or Twilio transports.
- [Google Chat plugin reference](https://docs.openclaw.ai/plugins/reference/googlechat): OpenClaw Google Chat channel plugin for spaces and direct messages.
- [Gradium plugin reference](https://docs.openclaw.ai/plugins/reference/gradium): Adds text-to-speech provider support.
- [Groq plugin reference](https://docs.openclaw.ai/plugins/reference/groq): Adds Groq model provider support to OpenClaw.
- [Hugging Face plugin reference](https://docs.openclaw.ai/plugins/reference/huggingface): Adds Hugging Face model provider support to OpenClaw.
- [Imap plugin reference](https://docs.openclaw.ai/plugins/reference/imap): Watch IMAP mailboxes and dispatch authenticated incoming email to isolated agent sessions.
- [iMessage plugin reference](https://docs.openclaw.ai/plugins/reference/imessage): OpenClaw iMessage channel plugin using imsg on a signed-in Mac.
- [Inworld plugin reference](https://docs.openclaw.ai/plugins/reference/inworld): Inworld streaming text-to-speech (MP3, OGG_OPUS, PCM telephony).
- [IRC plugin reference](https://docs.openclaw.ai/plugins/reference/irc): OpenClaw IRC channel plugin.
- [Kie plugin reference](https://docs.openclaw.ai/plugins/reference/kie): Adds Kie model provider support to OpenClaw.
- [Kilocode plugin reference](https://docs.openclaw.ai/plugins/reference/kilocode): Adds Kilocode model provider support to OpenClaw.
- [Kimi plugin reference](https://docs.openclaw.ai/plugins/reference/kimi): Adds Kimi, Kimi Code, Kimi Coding model provider support to OpenClaw.
- [LINE plugin reference](https://docs.openclaw.ai/plugins/reference/line): OpenClaw LINE channel plugin for LINE Bot API chats.
- [Linux Node plugin reference](https://docs.openclaw.ai/plugins/reference/linux-node): Desktop notifications, camera capture, and location for Linux node hosts.
- [LiteLLM plugin reference](https://docs.openclaw.ai/plugins/reference/litellm): Adds LiteLLM model provider support to OpenClaw.
- [Llama Cpp plugin reference](https://docs.openclaw.ai/plugins/reference/llama-cpp): Managed and external llama.cpp servers for GGUF chat and embeddings.
- [LLM Task plugin reference](https://docs.openclaw.ai/plugins/reference/llm-task): Generic JSON-only LLM tool for structured tasks callable from workflows.
- [LM Studio plugin reference](https://docs.openclaw.ai/plugins/reference/lmstudio): Adds LM Studio model provider support to OpenClaw.
- [Lobster plugin reference](https://docs.openclaw.ai/plugins/reference/lobster): Lobster workflow tool plugin for typed pipelines and resumable approvals.
- [Logbook plugin reference](https://docs.openclaw.ai/plugins/reference/logbook): Automatic work journal: captures periodic screen snapshots from a paired node and turns them into a reviewable timeline of your day.
- [LongCat plugin reference](https://docs.openclaw.ai/plugins/reference/longcat): OpenClaw LongCat provider plugin.
- [Matrix plugin reference](https://docs.openclaw.ai/plugins/reference/matrix): OpenClaw Matrix channel plugin for rooms and direct messages.
- [Mattermost plugin reference](https://docs.openclaw.ai/plugins/reference/mattermost): OpenClaw Mattermost channel plugin.
- [Memory Core plugin reference](https://docs.openclaw.ai/plugins/reference/memory-core): Adds agent-callable tools.
- [Memory Lancedb plugin reference](https://docs.openclaw.ai/plugins/reference/memory-lancedb): OpenClaw LanceDB-backed long-term memory plugin with auto-recall, auto-capture, and vector search.
- [Memory Wiki plugin reference](https://docs.openclaw.ai/plugins/reference/memory-wiki): Persistent wiki compiler and Obsidian-friendly knowledge vault for OpenClaw.
- [Meta plugin reference](https://docs.openclaw.ai/plugins/reference/meta): Adds Meta model provider support to OpenClaw.
- [Microsoft plugin reference](https://docs.openclaw.ai/plugins/reference/microsoft): Adds text-to-speech provider support.
- [Microsoft Foundry plugin reference](https://docs.openclaw.ai/plugins/reference/microsoft-foundry): Adds Microsoft Foundry model provider support to OpenClaw.
- [Migrate Claude plugin reference](https://docs.openclaw.ai/plugins/reference/migrate-claude): Imports Claude Code and Claude Desktop instructions, MCP servers, skills, and safe configuration into OpenClaw.
- [Migrate Hermes plugin reference](https://docs.openclaw.ai/plugins/reference/migrate-hermes): Imports Hermes configuration, memories, skills, and supported credentials into OpenClaw.
- [MiniMax plugin reference](https://docs.openclaw.ai/plugins/reference/minimax): Adds MiniMax, MiniMax Portal model provider support to OpenClaw.
- [Mistral plugin reference](https://docs.openclaw.ai/plugins/reference/mistral): Adds Mistral model provider support to OpenClaw.
- [Moonshot plugin reference](https://docs.openclaw.ai/plugins/reference/moonshot): Adds Moonshot model provider support to OpenClaw.
- [Microsoft Teams plugin reference](https://docs.openclaw.ai/plugins/reference/msteams): OpenClaw Microsoft Teams channel plugin for bot conversations.
- [Mxc plugin reference](https://docs.openclaw.ai/plugins/reference/mxc): OS-level sandboxed tool execution via MXC: runs commands in a Windows ProcessContainer with configured MXC policy files.
- [Nextcloud Talk plugin reference](https://docs.openclaw.ai/plugins/reference/nextcloud-talk): OpenClaw Nextcloud Talk channel plugin for conversations.
- [Nostr plugin reference](https://docs.openclaw.ai/plugins/reference/nostr): OpenClaw Nostr channel plugin for NIP-04 encrypted direct messages.
- [Novita plugin reference](https://docs.openclaw.ai/plugins/reference/novita): Adds Novita, Novita AI, Novitaai model provider support to OpenClaw.
- [NVIDIA plugin reference](https://docs.openclaw.ai/plugins/reference/nvidia): Adds NVIDIA model provider support to OpenClaw.
- [Oc Path plugin reference](https://docs.openclaw.ai/plugins/reference/oc-path): Adds the openclaw path CLI for oc:// workspace file addressing.
- [Ollama plugin reference](https://docs.openclaw.ai/plugins/reference/ollama): Adds Ollama, Ollama Cloud model provider support to OpenClaw.
- [Onepassword plugin reference](https://docs.openclaw.ai/plugins/reference/onepassword): 1Password SecretRef resolver and curated agent broker with approval policy and SQLite audit history.
- [Onnx plugin reference](https://docs.openclaw.ai/plugins/reference/onnx): Local typed decisions using pinned ONNX classifiers.
- [OpenAI plugin reference](https://docs.openclaw.ai/plugins/reference/openai): Adds OpenAI model provider support to OpenClaw.
- [OpenCode plugin reference](https://docs.openclaw.ai/plugins/reference/opencode): Adds OpenCode model provider support to OpenClaw.
- [OpenCode Go plugin reference](https://docs.openclaw.ai/plugins/reference/opencode-go): Adds OpenCode Go model provider support to OpenClaw.
- [OpenRouter plugin reference](https://docs.openclaw.ai/plugins/reference/openrouter): Adds OpenRouter model provider support to OpenClaw.
- [Openshell plugin reference](https://docs.openclaw.ai/plugins/reference/openshell): OpenClaw sandbox backend for the NVIDIA OpenShell CLI with mirrored local workspaces and SSH command execution.
- [Perplexity plugin reference](https://docs.openclaw.ai/plugins/reference/perplexity): Adds web search provider support.
- [PixVerse plugin reference](https://docs.openclaw.ai/plugins/reference/pixverse): OpenClaw PixVerse video generation provider plugin.
- [Policy plugin reference](https://docs.openclaw.ai/plugins/reference/policy): Adds policy-backed doctor checks for workspace conformance.
- [QA Channel plugin reference](https://docs.openclaw.ai/plugins/reference/qa-channel): OpenClaw QA synthetic channel plugin.
- [QA Lab plugin reference](https://docs.openclaw.ai/plugins/reference/qa-lab): OpenClaw QA lab plugin with private debugger UI and scenario runner.
- [Qianfan plugin reference](https://docs.openclaw.ai/plugins/reference/qianfan): Adds Qianfan model provider support to OpenClaw.
- [QQ Bot plugin reference](https://docs.openclaw.ai/plugins/reference/qqbot): OpenClaw QQ Bot channel plugin for group and direct-message workflows.
- [Qwen plugin reference](https://docs.openclaw.ai/plugins/reference/qwen): Adds Qwen, Qwen Cloud, Model Studio, DashScope, Qwen Token Plan, Bailian Token Plan model provider support to OpenClaw.
- [Radius plugin reference](https://docs.openclaw.ai/plugins/reference/radius): Radius model gateway provider.
- [Raft plugin reference](https://docs.openclaw.ai/plugins/reference/raft): OpenClaw Raft channel plugin for secure CLI wake bridges.
- [Reef plugin reference](https://docs.openclaw.ai/plugins/reference/reef): Guarded end-to-end encrypted claw channel.
- [Runway plugin reference](https://docs.openclaw.ai/plugins/reference/runway): Adds video generation provider support.
- [SearXNG plugin reference](https://docs.openclaw.ai/plugins/reference/searxng): Adds web search provider support.
- [Senseaudio plugin reference](https://docs.openclaw.ai/plugins/reference/senseaudio): Adds media understanding provider support.
- [Session Share plugin reference](https://docs.openclaw.ai/plugins/reference/session-share): Read-only OpenClaw sessions on paired gateways.
- [SGLang plugin reference](https://docs.openclaw.ai/plugins/reference/sglang): Adds SGLang model provider support to OpenClaw.
- [Signal plugin reference](https://docs.openclaw.ai/plugins/reference/signal): OpenClaw Signal channel plugin.
- [Slack plugin reference](https://docs.openclaw.ai/plugins/reference/slack): OpenClaw Slack channel plugin for channels, DMs, commands, and app events.
- [Slack huddles plugin reference](https://docs.openclaw.ai/plugins/reference/slack-huddles): Join Slack huddles through a dedicated Slack user in Chrome.
- [Sms plugin reference](https://docs.openclaw.ai/plugins/reference/sms): Twilio SMS/MMS channel plugin for OpenClaw messages.
- [StepFun plugin reference](https://docs.openclaw.ai/plugins/reference/stepfun): Adds StepFun, StepFun Plan model provider support to OpenClaw.
- [Synology Chat plugin reference](https://docs.openclaw.ai/plugins/reference/synology-chat): Synology Chat channel plugin for OpenClaw channels and direct messages.
- [Synthetic plugin reference](https://docs.openclaw.ai/plugins/reference/synthetic): Adds Synthetic model provider support to OpenClaw.
- [Talk Voice plugin reference](https://docs.openclaw.ai/plugins/reference/talk-voice): Manage Talk voice selection (list/set).
- [Tavily plugin reference](https://docs.openclaw.ai/plugins/reference/tavily): Adds agent-callable tools. Adds web search provider support.
- [Team Reports plugin reference](https://docs.openclaw.ai/plugins/reference/team-reports): Daily, weekly, and monthly team activity reports from GitHub and Discord, with model-written summaries, served in the Control UI.
- [Microsoft Teams meetings plugin reference](https://docs.openclaw.ai/plugins/reference/teams-meetings): Join Microsoft Teams meetings as a Chrome browser guest.
- [Telegram plugin reference](https://docs.openclaw.ai/plugins/reference/telegram): OpenClaw Telegram channel plugin.
- [Tencent plugin reference](https://docs.openclaw.ai/plugins/reference/tencent): Adds Tencent TokenHub, Tencent Tokenplan model provider support to OpenClaw.
- [Tlon plugin reference](https://docs.openclaw.ai/plugins/reference/tlon): OpenClaw Tlon/Urbit channel plugin for chat workflows.
- [Together plugin reference](https://docs.openclaw.ai/plugins/reference/together): Adds Together model provider support to OpenClaw.
- [Tokenjuice plugin reference](https://docs.openclaw.ai/plugins/reference/tokenjuice): Compacts exec and bash tool results with tokenjuice reducers.
- [TTS Local CLI plugin reference](https://docs.openclaw.ai/plugins/reference/tts-local-cli): Adds text-to-speech provider support.
- [Twitch plugin reference](https://docs.openclaw.ai/plugins/reference/twitch): OpenClaw Twitch channel plugin for chat and moderation workflows.
- [Typesafe plugin reference](https://docs.openclaw.ai/plugins/reference/typesafe): Typed decision provider for hosted Jev and local System One models.
- [Vault plugin reference](https://docs.openclaw.ai/plugins/reference/vault): HashiCorp Vault SecretRef provider integration.
- [Venice plugin reference](https://docs.openclaw.ai/plugins/reference/venice): Adds Venice model provider support to OpenClaw.
- [Vercel AI Gateway plugin reference](https://docs.openclaw.ai/plugins/reference/vercel-ai-gateway): Adds Vercel AI Gateway model provider support to OpenClaw.
- [Visitor Access plugin reference](https://docs.openclaw.ai/plugins/reference/visitor-access): Manage expiring visitor grants through one Cloudflare Access email policy.
- [vLLM plugin reference](https://docs.openclaw.ai/plugins/reference/vllm): Adds vLLM model provider support to OpenClaw.
- [Voice Call plugin reference](https://docs.openclaw.ai/plugins/reference/voice-call): OpenClaw voice-call plugin for Twilio, Telnyx, and Plivo phone calls.
- [Volcengine plugin reference](https://docs.openclaw.ai/plugins/reference/volcengine): Adds Volcengine, Volcengine Plan model provider support to OpenClaw.
- [Voyage plugin reference](https://docs.openclaw.ai/plugins/reference/voyage): Adds embedding provider support, including memory search.
- [Vydra plugin reference](https://docs.openclaw.ai/plugins/reference/vydra): Adds Vydra model provider support to OpenClaw.
- [Web Readability plugin reference](https://docs.openclaw.ai/plugins/reference/web-readability): Extract readable article content from local HTML web fetch responses.
- [WhatsApp plugin reference](https://docs.openclaw.ai/plugins/reference/whatsapp): OpenClaw WhatsApp channel plugin for WhatsApp Web chats.
- [Workboard plugin reference](https://docs.openclaw.ai/plugins/reference/workboard): Dashboard workboard for agent-owned issues and sessions.
- [X plugin reference](https://docs.openclaw.ai/plugins/reference/x): Allowlisted X mentions and public replies.
- [xAI plugin reference](https://docs.openclaw.ai/plugins/reference/xai): xAI provider for Grok models, Grok web and X search, code execution, speech, and media generation.
- [Xiaomi plugin reference](https://docs.openclaw.ai/plugins/reference/xiaomi): Adds Xiaomi, Xiaomi Token Plan model provider support to OpenClaw.
- [Z.AI plugin reference](https://docs.openclaw.ai/plugins/reference/zai): Adds Z.AI model provider support to OpenClaw.
- [Zalo plugin reference](https://docs.openclaw.ai/plugins/reference/zalo): OpenClaw Zalo channel plugin for bot and webhook chats.
- [Zalo Personal plugin reference](https://docs.openclaw.ai/plugins/reference/zalouser): OpenClaw Zalo Personal Account plugin via native zca-js integration.
- [Zoom meetings plugin reference](https://docs.openclaw.ai/plugins/reference/zoom-meetings): Join Zoom meetings as a Chrome browser guest.
- [Agent harness plugins](https://docs.openclaw.ai/plugins/sdk-agent-harness): Experimental SDK surface for plugins that replace the low level embedded agent executor
- [Agent harness attempt runtime](https://docs.openclaw.ai/plugins/sdk-agent-harness/attempt-runtime): Runtime helpers a selected harness calls during and after an attempt: injection, middleware, outcome classification, usage, and agent-end side effects
- [Agent harness core ownership](https://docs.openclaw.ai/plugins/sdk-agent-harness/core-ownership): What OpenClaw core prepares and owns before a harness runs an attempt, and the contracts a harness can declare to take some of it back
- [Agent harness native inventories](https://docs.openclaw.ai/plugins/sdk-agent-harness/native-inventories): Read-only model and MCP catalogs a harness reports from its own native runtime
- [Register an agent harness](https://docs.openclaw.ai/plugins/sdk-agent-harness/registration): Register an AgentHarnessV2, plus the optional isolated-completion and delegated-execution capabilities
- [Agent harness runtime configuration](https://docs.openclaw.ai/plugins/sdk-agent-harness/runtime-config): Operator config for the native Codex harness mode and for strict provider or model runtime policy
- [Agent harness selection policy](https://docs.openclaw.ai/plugins/sdk-agent-harness/selection-policy): How OpenClaw picks a harness after provider and model resolution, and why most harnesses also register a provider
- [Agent harness sessions and results](https://docs.openclaw.ai/plugins/sdk-agent-harness/sessions-and-results): Native session bindings, the OpenClaw transcript mirror, tool and media result delivery, terminal tool outcomes, and settled-turn finalization
- [Agent harness user input and execution authority](https://docs.openclaw.ai/plugins/sdk-agent-harness/user-input-and-execution): Blocking user-input surfaces, host tool capabilities, exec review outcomes, and paired-device command authority
- [Channel inbound API](https://docs.openclaw.ai/plugins/sdk-channel-inbound): Inbound event helpers for channel plugins: context building, shared runner orchestration, session record, and prepared reply dispatch
- [Channel ingress API](https://docs.openclaw.ai/plugins/sdk-channel-ingress): Experimental channel ingress API for inbound message authorization
- [Channel outbound API](https://docs.openclaw.ai/plugins/sdk-channel-outbound): Outbound message lifecycle API for channel plugins: adapters, receipts, durable sends, live preview, and reply pipeline helpers
- [Building channel plugins](https://docs.openclaw.ai/plugins/sdk-channel-plugins): Step-by-step guide to building a messaging channel plugin for OpenClaw
- [Channel approvals](https://docs.openclaw.ai/plugins/sdk-channel-plugins/approvals): Approval capabilities for channel plugins: approval auth, payload lifecycle and setup guidance, native approval delivery, and the narrower approval runtime subpaths
- [Durable channel ingress](https://docs.openclaw.ai/plugins/sdk-channel-plugins/durable-ingress): Durable inbound ingress for channel plugins: the ingress resolver, replay dedupe, transport retention classes, at-least-once side effects, and the reload and restart contract
- [Channel mention policy](https://docs.openclaw.ai/plugins/sdk-channel-plugins/mention-policy): Split inbound mention handling into plugin-owned evidence gathering and shared policy evaluation
- [Channel message adapter](https://docs.openclaw.ai/plugins/sdk-channel-plugins/message-adapter): The channel message adapter surface: live and finalizer capabilities, progress visibility, commentary delivery, and native TTS voice delivery
- [Channel sessions and bindings](https://docs.openclaw.ai/plugins/sdk-channel-plugins/sessions-and-bindings): Session conversation grammar, conversation route ownership, and account-scoped conversation binding support for channel plugins
- [Channel setup and config](https://docs.openclaw.ai/plugins/sdk-channel-plugins/setup-and-config): Setup subpaths, account schemas and inheritance, and the other narrow channel subpaths for config, inbound, targets, and threading
- [Channel status and media](https://docs.openclaw.ai/plugins/sdk-channel-plugins/status-and-media): Channel runtime status signals plus media and payload shaping: lifecycle status, typing indicators, media source params, and native payload shaping
- [Plugin entry points](https://docs.openclaw.ai/plugins/sdk-entrypoints): Reference for defineToolPlugin, definePluginEntry, defineChannelPluginEntry, and defineSetupPluginEntry
- [Plugin SDK defineChannelPluginEntry helper](https://docs.openclaw.ai/plugins/sdk-entrypoints/define-channel-plugin-entry): The channel entry helper, its mode-gated callbacks, and CLI registration
- [Plugin SDK definePluginEntry helper](https://docs.openclaw.ai/plugins/sdk-entrypoints/define-plugin-entry): The general entry helper, its fields, and the session catalog provider contract
- [Plugin SDK defineSetupPluginEntry helper](https://docs.openclaw.ai/plugins/sdk-entrypoints/define-setup-plugin-entry): The lightweight setup entry helper and the narrow setup helper families
- [Plugin SDK defineToolPlugin helper](https://docs.openclaw.ai/plugins/sdk-entrypoints/define-tool-plugin): The tool-only entry helper, its schemas, and generated manifest metadata
- [Plugin SDK native provider factories](https://docs.openclaw.ai/plugins/sdk-entrypoints/native-providers): Factory-form speech, transcription, and voice registration, and Computer Use providers
- [Plugin SDK package entries](https://docs.openclaw.ai/plugins/sdk-entrypoints/package-entries): package.json openclaw source and runtime entry fields and their resolution order
- [Plugin SDK registration mode](https://docs.openclaw.ai/plugins/sdk-entrypoints/registration-mode): How api.registrationMode reports the load mode and what to register in each
- [Plugin SDK tool policy and sandbox helpers](https://docs.openclaw.ai/plugins/sdk-entrypoints/tool-policy-and-sandbox): toolPolicy matchers, tool-group expansion, and splitSandboxBindSpec
- [Plugin SDK migration](https://docs.openclaw.ai/plugins/sdk-migration): Migrate from the legacy backwards-compatibility layer to the modern plugin SDK
- [Compatibility policy and records](https://docs.openclaw.ai/plugins/sdk-migration/compatibility-policy): The external-plugin compatibility policy and the dated per-surface compatibility records
- [How to migrate a plugin](https://docs.openclaw.ai/plugins/sdk-migration/how-to-migrate): Ordered steps for moving a plugin off the removed SDK compatibility layer
- [Import path reference](https://docs.openclaw.ai/plugins/sdk-migration/import-paths): Which typed-public SDK subpath replaces each legacy import, including the removed channel facades
- [Removal timeline](https://docs.openclaw.ai/plugins/sdk-migration/removal-timeline): When deprecated plugin SDK surfaces become eligible for removal
- [Removed surfaces and replacements](https://docs.openclaw.ai/plugins/sdk-migration/removed-surfaces): Removed SDK surfaces and the replacement for each removed or deprecated API
- [Talk and realtime voice migration](https://docs.openclaw.ai/plugins/sdk-migration/talk): Migrating realtime voice, telephony, and meeting code to the unified Talk session API
- [Plugin SDK overview](https://docs.openclaw.ai/plugins/sdk-overview): Import map, registration API reference, and SDK architecture
- [Plugin SDK capability registration](https://docs.openclaw.ai/plugins/sdk-overview/capabilities): Provider, storage, worker, and embedding registration on OpenClawPluginApi
- [Plugin SDK CLI and discovery registration](https://docs.openclaw.ai/plugins/sdk-overview/cli-and-discovery): Gateway discovery services, CLI registrars and descriptors, and CLI backend registration
- [Plugin SDK events and hook semantics](https://docs.openclaw.ai/plugins/sdk-overview/events-and-hooks): Typed lifecycle hooks and the decision rules each hook applies
- [Plugin SDK host hooks for workflow plugins](https://docs.openclaw.ai/plugins/sdk-overview/host-hooks): Session extensions, trusted tool policies, Control UI descriptors, and runtime lifecycle
- [Plugin SDK imports and module layout](https://docs.openclaw.ai/plugins/sdk-overview/imports): Which plugin SDK subpath to import from, and how to lay out a plugin's own barrels
- [Plugin SDK infrastructure registration](https://docs.openclaw.ai/plugins/sdk-overview/infrastructure): Hooks, HTTP routes, Gateway methods, services, and the webhook and SQLite helpers
- [Plugin SDK memory and context slots](https://docs.openclaw.ai/plugins/sdk-overview/memory-and-context): The exclusive context-engine and memory-capability slots and their embedding adapters
- [Plugin SDK tools and commands](https://docs.openclaw.ai/plugins/sdk-overview/tools-and-commands): Registering agent tools, custom commands, node-host commands, and widget presenters
- [Building provider plugins](https://docs.openclaw.ai/plugins/sdk-provider-plugins): Step-by-step guide to building a model provider plugin for OpenClaw
- [Provider hook families](https://docs.openclaw.ai/plugins/sdk-provider-plugins/hook-families): Shared replay, stream, and tool-compat family builders for provider plugins
- [Provider media and search](https://docs.openclaw.ai/plugins/sdk-provider-plugins/media-and-search): Embedding, image and video generation, web fetch, and web search provider capabilities
- [Provider model catalogs](https://docs.openclaw.ai/plugins/sdk-provider-plugins/model-catalogs): Live model discovery, catalog helpers, and provider-owned pricing normalization
- [Provider hook wiring](https://docs.openclaw.ai/plugins/sdk-provider-plugins/runtime-hooks): Per-hook provider wiring for auth exchange, headers, transport identity, usage, and the hook order table
- [Provider voice capabilities](https://docs.openclaw.ai/plugins/sdk-provider-plugins/voice-and-audio): Speech, realtime transcription, realtime voice, and media understanding provider capabilities
- [Plugin runtime helpers](https://docs.openclaw.ai/plugins/sdk-runtime): api.runtime -- the injected runtime helpers available to plugins
- [Plugin runtime agent helpers](https://docs.openclaw.ai/plugins/sdk-runtime/agent): Agent identity, directories, session store, transcripts, and sandbox authority
- [Plugin runtime background work](https://docs.openclaw.ai/plugins/sdk-runtime/background-work): Hook agent turns and subagent runs
- [Plugin runtime channel helpers](https://docs.openclaw.ai/plugins/sdk-runtime/channel): Channel-specific runtime helper groups for chunking, routing, pairing, media, and mentions
- [Plugin runtime config and utilities](https://docs.openclaw.ai/plugins/sdk-runtime/config-and-utilities): Runtime config reads and writes, plus the shared process, error, and model-picker utilities
- [Plugin runtime Gateway and nodes](https://docs.openclaw.ai/plugins/sdk-runtime/gateway-and-nodes): In-process Gateway requests, paired node invocation, and Gateway service events
- [Plugin runtime media helpers](https://docs.openclaw.ai/plugins/sdk-runtime/media): Speech, media understanding, image/video/music generation, web search, and media utilities
- [Plugin runtime model helpers](https://docs.openclaw.ai/plugins/sdk-runtime/models): Host-owned completions, model-selection policy, and provider auth resolution
- [Plugin runtime state and system](https://docs.openclaw.ai/plugins/sdk-runtime/state-and-system): Config snapshot, SQLite-backed plugin state, system utilities, events, and logging
- [Plugin setup and config](https://docs.openclaw.ai/plugins/sdk-setup): Setup wizards, setup-entry.ts, config schemas, and package.json metadata
- [Plugin SDK subpaths](https://docs.openclaw.ai/plugins/sdk-subpaths): Plugin SDK subpath catalog: which imports live where, grouped by area
- [Plugin testing](https://docs.openclaw.ai/plugins/sdk-testing): Testing utilities and patterns for OpenClaw plugins
- [Session Share plugin](https://docs.openclaw.ai/plugins/session-share): Share selected OpenClaw sessions read-only with a paired team Gateway
- [Slack huddles plugin](https://docs.openclaw.ai/plugins/slack-huddles): Slack huddles plugin: join active huddles through a signed-in Chrome user account
- [Team Reports plugin](https://docs.openclaw.ai/plugins/team-reports): Daily, weekly, and monthly team reports from GitHub activity and Discord discussion
- [Microsoft Teams meetings plugin](https://docs.openclaw.ai/plugins/teams-meetings): Microsoft Teams meetings plugin: join work or consumer meetings as a Chrome browser guest
- [Tool plugins](https://docs.openclaw.ai/plugins/tool-plugins): Build simple typed agent tools with defineToolPlugin and openclaw plugins init/build/validate
- [TypeSafe AI](https://docs.openclaw.ai/plugins/typesafe): Use hosted Jev or a local System One server for typed decisions
- [Vault SecretRefs](https://docs.openclaw.ai/plugins/vault): Use the bundled Vault plugin to resolve SecretRefs from HashiCorp Vault
- [Voice call plugin](https://docs.openclaw.ai/plugins/voice-call): Place outbound and accept inbound voice calls via Twilio, Telnyx, or Plivo, with optional realtime voice and streaming transcription
- [Voice call configuration](https://docs.openclaw.ai/plugins/voice-call/configuration): Plugin config keys, the call owner, the config reference table, and session scope
- [Voice call realtime and streaming](https://docs.openclaw.ai/plugins/voice-call/realtime-and-streaming): Realtime voice conversations, tool policy, agent voice context, and streaming transcription
- [Voice call security and interfaces](https://docs.openclaw.ai/plugins/voice-call/security-and-interfaces): Webhook security options and the CLI, agent tool, and Gateway RPC surfaces
- [Voice call troubleshooting](https://docs.openclaw.ai/plugins/voice-call/troubleshooting): Fixes for voice-call setup, webhooks, signatures, Meet dial-in, and silent calls
- [Voice call TTS and inbound calls](https://docs.openclaw.ai/plugins/voice-call/tts-and-inbound-calls): Telephony TTS, inbound policy, per-number routing, and call lifecycle timers
- [Workboard plugin](https://docs.openclaw.ai/plugins/workboard): Optional Workboard boards for agent-owned cards and rule-based session views
- [Zalo personal plugin](https://docs.openclaw.ai/plugins/zalouser): Zalo Personal plugin: QR login + messaging via native zca-js (plugin install + channel config + tool)
- [Zoom meetings plugin](https://docs.openclaw.ai/plugins/zoom-meetings): Zoom meetings plugin: join meetings as a Chrome browser guest
- [OpenProse removal and migration](https://docs.openclaw.ai/prose): OpenClaw no longer bundles OpenProse or the /prose command. Move to the maintained upstream Agent Skill and clean stale plugin configuration.
- [Provider directory](https://docs.openclaw.ai/providers): Model providers (LLMs) supported by OpenClaw
- [Alibaba Model Studio](https://docs.openclaw.ai/providers/alibaba): Alibaba Model Studio Wan video generation in OpenClaw
- [Anthropic](https://docs.openclaw.ai/providers/anthropic): Use Anthropic Claude via API keys or Claude CLI in OpenClaw
- [Arcee AI](https://docs.openclaw.ai/providers/arcee): Arcee AI setup (auth + model selection)
- [Azure Speech](https://docs.openclaw.ai/providers/azure-speech): Azure AI Speech text-to-speech for OpenClaw replies
- [Baseten](https://docs.openclaw.ai/providers/baseten): Baseten setup for DeepSeek V4.1 Flash and hosted Model APIs
- [Amazon Bedrock](https://docs.openclaw.ai/providers/bedrock): Use Amazon Bedrock (Converse API) models with OpenClaw
- [Amazon Bedrock Mantle](https://docs.openclaw.ai/providers/bedrock-mantle): Use Amazon Bedrock Mantle OpenAI-compatible and Claude Messages models with OpenClaw
- [Cerebras](https://docs.openclaw.ai/providers/cerebras): Cerebras setup (auth + model selection)
- [Chutes](https://docs.openclaw.ai/providers/chutes): Chutes setup (OAuth or API key, model discovery, aliases)
- [Claude Max API proxy](https://docs.openclaw.ai/providers/claude-max-api-proxy): Community proxy to expose Claude subscription credentials as an OpenAI-compatible endpoint
- [ClawRouter](https://docs.openclaw.ai/providers/clawrouter): Route credential-scoped models through ClawRouter and show managed quotas
- [Cloudflare AI gateway](https://docs.openclaw.ai/providers/cloudflare-ai-gateway): Cloudflare AI Gateway setup (auth + model selection)
- [Cohere](https://docs.openclaw.ai/providers/cohere): Cohere setup (auth + model selection)
- [ComfyUI](https://docs.openclaw.ai/providers/comfy): ComfyUI workflow generation and Comfy Cloud MCP OAuth setup in OpenClaw
- [Deepgram](https://docs.openclaw.ai/providers/deepgram): Deepgram transcription for inbound voice notes
- [DeepInfra](https://docs.openclaw.ai/providers/deepinfra): Use DeepInfra's unified API to access the most popular open source and frontier models in OpenClaw
- [DeepSeek](https://docs.openclaw.ai/providers/deepseek): DeepSeek setup (auth + model selection)
- [ds4](https://docs.openclaw.ai/providers/ds4): Run OpenClaw through ds4, a local DeepSeek V4 Flash OpenAI-compatible server
- [ElevenLabs](https://docs.openclaw.ai/providers/elevenlabs): Use ElevenLabs speech, Scribe STT, and realtime transcription with OpenClaw
- [Fal](https://docs.openclaw.ai/providers/fal): fal image, video, and music generation setup in OpenClaw
- [Featherless AI](https://docs.openclaw.ai/providers/featherless): Featherless AI setup, model selection, and tool calling
- [Fireworks](https://docs.openclaw.ai/providers/fireworks): Fireworks setup (auth + model selection)
- [Fish Audio](https://docs.openclaw.ai/providers/fish-audio): Use Fish Audio S2.1 hosted TTS or local S2 Pro on Apple silicon
- [GitHub Copilot](https://docs.openclaw.ai/providers/github-copilot): Sign in to GitHub Copilot from OpenClaw using the device flow or non-interactive token import
- [GMI Cloud](https://docs.openclaw.ai/providers/gmi): Use GMI Cloud's OpenAI-compatible API with OpenClaw
- [Google (Gemini)](https://docs.openclaw.ai/providers/google): Google Gemini setup (AI Studio API key, Vertex AI, optional CLI runtime, and multimodal tools)
- [Gradium](https://docs.openclaw.ai/providers/gradium): Use Gradium text-to-speech in OpenClaw
- [Groq](https://docs.openclaw.ai/providers/groq): Groq setup (auth + model selection + Whisper transcription)
- [Hugging Face (inference)](https://docs.openclaw.ai/providers/huggingface): Hugging Face Inference setup (auth + model selection)
- [Inworld](https://docs.openclaw.ai/providers/inworld): Inworld streaming text-to-speech for OpenClaw replies
- [Kie AI](https://docs.openclaw.ai/providers/kie): Kie AI video generation setup and supported models
- [Kilo Gateway](https://docs.openclaw.ai/providers/kilocode): Use Kilo Gateway's unified API to access many models in OpenClaw
- [LiteLLM](https://docs.openclaw.ai/providers/litellm): Run OpenClaw through LiteLLM Proxy for unified model access and cost tracking
- [llmman](https://docs.openclaw.ai/providers/llmman): Run OpenClaw with llmman (local models, hosted providers, and hybrid local + hosted routing)
- [LM Studio](https://docs.openclaw.ai/providers/lmstudio): Run OpenClaw with LM Studio
- [LongCat](https://docs.openclaw.ai/providers/longcat): LongCat API setup for LongCat-2.0
- [Meta](https://docs.openclaw.ai/providers/meta): Meta setup, authentication, and Muse Spark model selection
- [MiniMax](https://docs.openclaw.ai/providers/minimax): Use MiniMax models in OpenClaw
- [Mistral](https://docs.openclaw.ai/providers/mistral): Use Mistral models and Voxtral transcription with OpenClaw
- [Model provider quickstart](https://docs.openclaw.ai/providers/models): Model providers (LLMs) supported by OpenClaw
- [Moonshot AI](https://docs.openclaw.ai/providers/moonshot): Configure Moonshot Kimi models vs Kimi Coding (separate providers + keys)
- [NovitaAI](https://docs.openclaw.ai/providers/novita): Use NovitaAI's OpenAI-compatible API with OpenClaw
- [NVIDIA](https://docs.openclaw.ai/providers/nvidia): Use NVIDIA's OpenAI-compatible API in OpenClaw
- [Ollama](https://docs.openclaw.ai/providers/ollama): Run OpenClaw with Ollama (cloud and local models)
- [Ollama Cloud](https://docs.openclaw.ai/providers/ollama-cloud): Use Ollama Cloud directly with OpenClaw
- [Ollama advanced configuration](https://docs.openclaw.ai/providers/ollama/advanced): Context windows, thinking control, costs, embeddings, and streaming for Ollama
- [Ollama configuration](https://docs.openclaw.ai/providers/ollama/configuration): The basic, explicit, and custom base URL config shapes for the Ollama provider
- [Ollama model discovery](https://docs.openclaw.ai/providers/ollama/model-discovery): How OpenClaw discovers Ollama models implicitly, plus narrow smoke tests
- [Ollama node-local inference](https://docs.openclaw.ai/providers/ollama/node-local-inference): Delegate a bounded prompt to an Ollama model running on a paired node
- [Ollama config recipes](https://docs.openclaw.ai/providers/ollama/recipes): Copyable Ollama config recipes, model selection, and quick verification
- [Ollama setup](https://docs.openclaw.ai/providers/ollama/setup): Connect OpenClaw to Ollama: auth rules, onboarding, and cloud models through a local host
- [Ollama troubleshooting](https://docs.openclaw.ai/providers/ollama/troubleshooting): Fixes for common Ollama failures in OpenClaw
- [Ollama vision and image description](https://docs.openclaw.ai/providers/ollama/vision): Route image understanding through local or hosted Ollama vision models
- [Ollama Web Search](https://docs.openclaw.ai/providers/ollama/web-search): Use Ollama as the OpenClaw web_search provider
- [OpenAI](https://docs.openclaw.ai/providers/openai): Use OpenAI via API keys or Codex subscription in OpenClaw
- [OpenAI advanced configuration](https://docs.openclaw.ai/providers/openai/advanced): Prompt contribution, transport, Fast mode, compaction, and route compat
- [OpenAI authentication](https://docs.openclaw.ai/providers/openai/authentication): Choose Codex login, Sign in with ChatGPT (Beta), or an API key for the OpenAI capabilities you need
- [Azure OpenAI endpoints](https://docs.openclaw.ai/providers/openai/azure): Point the bundled openai provider at an Azure OpenAI resource
- [OpenAI coverage and cost](https://docs.openclaw.ai/providers/openai/coverage-and-cost): Which OpenAI capabilities OpenClaw exposes, embeddings, and spend reporting
- [OpenAI image generation](https://docs.openclaw.ai/providers/openai/image-and-video): Generate and edit images with OpenAI gpt-image
- [OpenAI models](https://docs.openclaw.ai/providers/openai/models): Pick an OpenAI model ref, including GPT-6.1 Sol, GPT-6 Astra, Sol, Luna, and the GPT-5.6 tiers
- [OpenAI runtimes and Codex auth](https://docs.openclaw.ai/providers/openai/runtimes): Choose an OpenAI runtime and understand native Codex auth
- [OpenAI setup](https://docs.openclaw.ai/providers/openai/setup): Connect OpenAI with an API key, Codex subscription, or Sign in with ChatGPT (Beta)
- [OpenAI voice and speech](https://docs.openclaw.ai/providers/openai/voice-and-speech): OpenAI text-to-speech, transcription, and realtime voice settings and auth
- [OpenCode](https://docs.openclaw.ai/providers/opencode): Use OpenCode Zen and Go catalogs with OpenClaw
- [OpenCode Go](https://docs.openclaw.ai/providers/opencode-go): Use the OpenCode Go catalog with the shared OpenCode setup
- [OpenRouter](https://docs.openclaw.ai/providers/openrouter): Use OpenRouter's unified API to access many models in OpenClaw
- [Perplexity](https://docs.openclaw.ai/providers/perplexity-provider): Perplexity web search provider setup (API key, search modes, filtering)
- [PixVerse](https://docs.openclaw.ai/providers/pixverse): PixVerse video generation setup in OpenClaw
- [Qianfan](https://docs.openclaw.ai/providers/qianfan): Use Qianfan's unified API to access many models in OpenClaw
- [Qwen](https://docs.openclaw.ai/providers/qwen): Use Qwen Cloud through its OpenClaw plugin
- [Radius](https://docs.openclaw.ai/providers/radius): Connect OpenClaw to Radius with browser sign-in or an organization API key
- [Runway](https://docs.openclaw.ai/providers/runway): Runway video generation setup in OpenClaw
- [SenseAudio](https://docs.openclaw.ai/providers/senseaudio): SenseAudio batch speech-to-text for inbound voice notes
- [SGLang](https://docs.openclaw.ai/providers/sglang): Run OpenClaw with SGLang (OpenAI-compatible self-hosted server)
- [StepFun](https://docs.openclaw.ai/providers/stepfun): Use StepFun models with OpenClaw
- [Synthetic](https://docs.openclaw.ai/providers/synthetic): Use Synthetic's Anthropic-compatible API in OpenClaw
- [Telnyx](https://docs.openclaw.ai/providers/telnyx): Telnyx setup for OpenAI-compatible AI inference
- [Tencent Cloud (TokenHub / TokenPlan)](https://docs.openclaw.ai/providers/tencent): Tencent Cloud TokenHub and TokenPlan setup for hy4-preview
- [Together AI](https://docs.openclaw.ai/providers/together): Together AI setup (auth + model selection)
- [Venice AI](https://docs.openclaw.ai/providers/venice): Use Venice AI privacy-focused models in OpenClaw
- [Vercel AI gateway](https://docs.openclaw.ai/providers/vercel-ai-gateway): Vercel AI Gateway setup (auth + model selection)
- [vLLM](https://docs.openclaw.ai/providers/vllm): Run OpenClaw with vLLM (OpenAI-compatible local server)
- [Volcengine (Doubao)](https://docs.openclaw.ai/providers/volcengine): Volcano Engine setup (Doubao models, coding endpoints, and Seed Speech TTS)
- [Vydra](https://docs.openclaw.ai/providers/vydra): Use Vydra image, video, and speech in OpenClaw
- [xAI](https://docs.openclaw.ai/providers/xai): Use xAI Grok models in OpenClaw
- [Xiaomi MiMo](https://docs.openclaw.ai/providers/xiaomi): Use Xiaomi MiMo pay-as-you-go and Token Plan models with OpenClaw
- [Z.AI](https://docs.openclaw.ai/providers/zai): Use Z.AI (GLM models) with OpenClaw
- [Default AGENTS.md](https://docs.openclaw.ai/reference/AGENTS.default): Default OpenClaw agent instructions and skills roster for the personal assistant setup
- [API usage and costs](https://docs.openclaw.ai/reference/api-usage-costs): Audit what can spend money, which keys are used, and how to view usage
- [Credits](https://docs.openclaw.ai/reference/credits): Project origin, contributors, and license.
- [Database schemas](https://docs.openclaw.ai/reference/database-schemas): OpenClaw SQLite database locations, schema versions, integrity checks, and downgrade recovery
- [Agent schema history](https://docs.openclaw.ai/reference/database-schemas/agent-schema-history): Per-agent database schema versions, their changes, and their first releases
- [Integrity, troubleshooting, and recovery](https://docs.openclaw.ai/reference/database-schemas/integrity-and-recovery): Integrity checks, common database errors, and the supported downgrade recovery path
- [Database layout](https://docs.openclaw.ai/reference/database-schemas/layout): Which SQLite database holds what, and the tables behind individual features
- [Per-person and companion storage](https://docs.openclaw.ai/reference/database-schemas/personal-data): Where personal GitHub connections, personal model accounts, and Apple companion journals live
- [State schema history](https://docs.openclaw.ai/reference/database-schemas/state-schema-history): Shared state database schema versions, their changes, and their first releases
- [Storage changes and release preflight](https://docs.openclaw.ai/reference/database-schemas/storage-changes): Proposing a persistent-store change, the review checkpoint, and preflighting a target release
- [Versioning contract](https://docs.openclaw.ai/reference/database-schemas/versioning): How OpenClaw records schema versions, when a bump is required, and how updaters cross one
- [Database access in workers](https://docs.openclaw.ai/reference/database-schemas/worker-access): Move Gateway database access into existing workers while preserving admission, revisions, and publication
- [Database worker migration inventory](https://docs.openclaw.ai/reference/database-schemas/worker-access-inventory): Generated inventory and migration priorities for synchronous SQLite access
- [Device model database](https://docs.openclaw.ai/reference/device-models): How OpenClaw vendors Apple device model identifiers for friendly names in the macOS app.
- [Full release validation](https://docs.openclaw.ai/reference/full-release-validation): Index of the Full Release Validation reference, one page per reader job
- [Continue a failed validation](https://docs.openclaw.ai/reference/full-release-validation/continuation): Continuing a failed Full Release Validation parent, attempt adoption rules, and the post-merge continuation proof
- [Dispatch a validation run](https://docs.openclaw.ai/reference/full-release-validation/dispatch): Dispatching Full Release Validation: Code SHA and Tooling SHA selection, helper inputs, and the immutable execution plan
- [Evidence to keep](https://docs.openclaw.ai/reference/full-release-validation/evidence): Which Full Release Validation evidence to retain, blocking lane outcomes, and the backing workflow files
- [Extended-stable and changelog-only validation](https://docs.openclaw.ai/reference/full-release-validation/extended-stable): Extended-stable validation dispatch, changelog-only Release SHA reuse, coverage policies, and Telegram deferrals and waivers
- [Release profiles and focused reruns](https://docs.openclaw.ai/reference/full-release-validation/profiles): Release profile coverage, full-only additions, and the focused rerun handles and suite filters
- [Release checks stages](https://docs.openclaw.ai/reference/full-release-validation/release-checks): OpenClaw Release Checks stage matrix and the Docker release-path chunk list
- [Top-level stages](https://docs.openclaw.ai/reference/full-release-validation/stages): The Full Release Validation umbrella stage matrix, evidence reuse, artifact producers, and decision states
- [Memory configuration reference](https://docs.openclaw.ai/reference/memory-config): Built-in memory search, admission exclusions, and dreaming configuration
- [@openclaw/ai package](https://docs.openclaw.ai/reference/openclaw-ai): The @openclaw/ai npm package: reusable model transports, isolated runtimes, and host policy ports
- [Prompt caching](https://docs.openclaw.ai/reference/prompt-caching): Prompt caching knobs, merge order, provider behavior, and tuning patterns
- [Pull request review flow](https://docs.openclaw.ai/reference/pull-request-review-flow): How Barnacle and ClawSweeper feedback helps move OpenClaw pull requests through review.
- [Release performance sweep](https://docs.openclaw.ai/reference/release-performance-sweep): Visual summary and technical evidence for the May 2026 performance, package-size, dependency, and shrinkwrap cleanup
- [Release policy](https://docs.openclaw.ai/reference/RELEASING): OpenClaw release channels, version numbers, validation, and published assets
- [Rich output protocol](https://docs.openclaw.ai/reference/rich-output-protocol): Rich output protocol for structured media, embeds, audio hints, and replies
- [RPC adapters](https://docs.openclaw.ai/reference/rpc): RPC adapters for external CLIs (signal-cli, imsg) and gateway patterns
- [Secret Placeholder Conventions](https://docs.openclaw.ai/reference/secret-placeholder-conventions): Secret-scanner-safe placeholder conventions for docs and examples
- [SecretRef credential surface](https://docs.openclaw.ai/reference/secretref-credential-surface): Canonical supported vs unsupported SecretRef credential surface
- [Session management deep dive](https://docs.openclaw.ai/reference/session-management-compaction): Deep dive: session store + transcripts, lifecycle, and (auto)compaction internals
- [Compaction behavior and settings](https://docs.openclaw.ai/reference/session-management-compaction/compaction): What compaction does, when auto-compaction runs, its settings, pluggable providers, and where it surfaces
- [Silent turns and the memory flush](https://docs.openclaw.ai/reference/session-management-compaction/housekeeping): The NO_REPLY silent-turn contract and the pre-compaction memory flush configuration
- [Store maintenance and retention](https://docs.openclaw.ai/reference/session-management-compaction/maintenance): The session.maintenance keys, disk-budget cleanup tiers, cron run retention, and the SQLite downgrade path
- [Session transcript working sets](https://docs.openclaw.ai/reference/session-management-compaction/resident-history): Proposed worker-owned transcript acquisition and bounded SessionManager working sets
- [Session keys, ids, and transcript events](https://docs.openclaw.ai/reference/session-management-compaction/schema): sessionKey patterns, sessionId lifecycle, SessionEntry fields, and the transcript event stream
- [Session state on disk](https://docs.openclaw.ai/reference/session-management-compaction/store): Where the Gateway keeps session rows and transcripts, and the on-disk paths per agent
- [Team immutable update design](https://docs.openclaw.ai/reference/team-immutable-update-design): Immutable update design: installation, preparation, native activation, and retained recovery contracts
- [AGENTS.md template](https://docs.openclaw.ai/reference/templates/AGENTS): Workspace template for AGENTS.md
- [AGENTS.dev template](https://docs.openclaw.ai/reference/templates/AGENTS.dev): Dev agent AGENTS.md (C-3PO)
- [BOOT.md template](https://docs.openclaw.ai/reference/templates/BOOT): Workspace template for BOOT.md
- [BOOTSTRAP.md template](https://docs.openclaw.ai/reference/templates/BOOTSTRAP): First-run ritual for new agents
- [Retired HEARTBEAT.md workspace file](https://docs.openclaw.ai/reference/templates/HEARTBEAT): Migration guide for the retired HEARTBEAT.md workspace file
- [IDENTITY template](https://docs.openclaw.ai/reference/templates/IDENTITY): Agent identity record
- [IDENTITY.dev template](https://docs.openclaw.ai/reference/templates/IDENTITY.dev): Dev agent identity (C-3PO)
- [Chief of staff soul](https://docs.openclaw.ai/reference/templates/roles/coordinator/CLAW)
- [Chief of staff operating program](https://docs.openclaw.ai/reference/templates/roles/coordinator/workspace/AGENTS)
- [Researcher soul](https://docs.openclaw.ai/reference/templates/roles/researcher/CLAW)
- [Researcher operating program](https://docs.openclaw.ai/reference/templates/roles/researcher/workspace/AGENTS)
- [Reviewer soul](https://docs.openclaw.ai/reference/templates/roles/reviewer/CLAW)
- [Reviewer operating program](https://docs.openclaw.ai/reference/templates/roles/reviewer/workspace/AGENTS)
- [Writer soul](https://docs.openclaw.ai/reference/templates/roles/writer/CLAW)
- [Writer operating program](https://docs.openclaw.ai/reference/templates/roles/writer/workspace/AGENTS)
- [SOUL.md template](https://docs.openclaw.ai/reference/templates/SOUL): Workspace template for SOUL.md
- [SOUL.dev template](https://docs.openclaw.ai/reference/templates/SOUL.dev): Dev agent soul (C-3PO)
- [TOOLS.md retired](https://docs.openclaw.ai/reference/templates/TOOLS): Retired TOOLS.md workspace template
- [USER template](https://docs.openclaw.ai/reference/templates/USER): Durable user preference and profile directives
- [USER.dev template](https://docs.openclaw.ai/reference/templates/USER.dev): Dev agent user profile (C-3PO)
- [Tests](https://docs.openclaw.ai/reference/test): Index of the OpenClaw testing reference, one page per reader job
- [Docker test suites](https://docs.openclaw.ai/reference/test/docker): The full Docker suite scheduler, its lanes, and the onboarding and QR smokes
- [Control UI, TUI, and E2E lanes](https://docs.openclaw.ai/reference/test/lanes): Control UI, TUI, extension, Gateway, and live lane commands and fixture rules
- [Run tests locally](https://docs.openclaw.ai/reference/test/local): Routine local test order, the core test commands, and the local PR gate
- [Test performance and benchmarks](https://docs.openclaw.ai/reference/test/performance): Import profiling, CPU and heap profiles, shard timings, and benchmark scripts
- [Remote test proof](https://docs.openclaw.ai/reference/test/remote-proof): When agents use Crabbox or Testbox, and the wrapper, lease, and trust rules
- [Test runner internals](https://docs.openclaw.ai/reference/test/runner-internals): Shared build locks, isolated test state and homes, and JSON report merging
- [Token use and costs](https://docs.openclaw.ai/reference/token-use): How OpenClaw builds prompt context and reports token usage + costs
- [Transcript hygiene](https://docs.openclaw.ai/reference/transcript-hygiene): Reference: provider-specific transcript sanitization and repair rules
- [Release notes](https://docs.openclaw.ai/releases): OpenClaw release notes, complete Markdown changelogs, and contribution records.
- [v2026.6.11](https://docs.openclaw.ai/releases/2026.6.11): Reliability fixes for replies, sends, reconnects, model setup, sessions, and safer admin defaults.
- [v2026.7.1](https://docs.openclaw.ai/releases/2026.7.1): Major Control UI and onboarding overhauls, big iOS, Android, and macOS updates, expanded model support, and substantial Telegram, Slack, Discord, and Apple Messages improvements.
- [v2026.8.1 (AKA OpenClaw 2.0)](https://docs.openclaw.ai/releases/2026.8.1): A rebuilt web experience, simpler onboarding, stronger memory and session continuity, and a very large reliability pass across OpenClaw.
- [v2026.8.1: Automations and Scheduling](https://docs.openclaw.ai/releases/2026.8.1/automations-and-scheduling)
- [v2026.8.1: Browser and Computer Use](https://docs.openclaw.ai/releases/2026.8.1/browser-and-computer-use)
- [v2026.8.1: Installation and Onboarding](https://docs.openclaw.ai/releases/2026.8.1/installation-and-onboarding)
- [v2026.8.1: Maintainer and Internal Changes](https://docs.openclaw.ai/releases/2026.8.1/maintainer-and-internal-changes)
- [v2026.8.1: Maintenance changes (part 1 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-1)
- [v2026.8.1: Maintenance changes (part 2 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-2)
- [v2026.8.1: Maintenance changes (part 3 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-3)
- [v2026.8.1: Maintenance changes (part 4 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-4)
- [v2026.8.1: Maintenance changes (part 5 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-5)
- [v2026.8.1: Maintenance changes (part 6 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-6)
- [v2026.8.1: Maintenance changes (part 7 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-7)
- [v2026.8.1: Maintenance changes (part 8 of 8)](https://docs.openclaw.ai/releases/2026.8.1/maintenance-changes-part-8)
- [v2026.8.1: Memory](https://docs.openclaw.ai/releases/2026.8.1/memory)
- [v2026.8.1: Messaging](https://docs.openclaw.ai/releases/2026.8.1/messaging)
- [v2026.8.1: Models and Providers](https://docs.openclaw.ai/releases/2026.8.1/models-and-providers)
- [v2026.8.1: Native Apps](https://docs.openclaw.ai/releases/2026.8.1/native-apps)
- [v2026.8.1: Other Bug Fixes](https://docs.openclaw.ai/releases/2026.8.1/other-bug-fixes)
- [v2026.8.1: Plugins and Integrations](https://docs.openclaw.ai/releases/2026.8.1/plugins-and-integrations)
- [v2026.8.1: Quality-of-Life Improvements](https://docs.openclaw.ai/releases/2026.8.1/quality-of-life-improvements)
- [v2026.8.1: Security and Privacy](https://docs.openclaw.ai/releases/2026.8.1/security-and-privacy)
- [v2026.8.1: Skills](https://docs.openclaw.ai/releases/2026.8.1/skills)
- [v2026.8.1: The New Web UI](https://docs.openclaw.ai/releases/2026.8.1/the-new-web-ui)
- [v2026.8.1: Updates and Maintenance](https://docs.openclaw.ai/releases/2026.8.1/updates-and-maintenance)
- [v2026.8.2](https://docs.openclaw.ai/releases/2026.8.2): Home beside your work, a Linux desktop companion, background sessions, browser control without a running Gateway, four new Control UI looks, and focused reply, voice, update, and recovery fixes.
- [v2026.9.1](https://docs.openclaw.ai/releases/2026.9.1): Mermaid diagrams across chat surfaces, a fuller Android experience, safer update recovery, and lower overhead for long conversations and large installations.
- [v2026.9.2](https://docs.openclaw.ai/releases/2026.9.2)
- [v2026.9.3](https://docs.openclaw.ai/releases/2026.9.3): Clean update recovery, faster session reconnects, live browser automation, revocable chat links, searchable meeting transcripts, and repository-backed cloud work.
- [v2026.9.4](https://docs.openclaw.ai/releases/2026.9.4): Plugin and skill discovery, visible skill learning, cloud-worker controls, GPT Image 2.5, and terminal questions.
- [v2026.9.5](https://docs.openclaw.ai/releases/2026.9.5): Atomic Updates, plugin hot reloading, conversation sharing, GPT Live, shared browser pages, conversation archiving, and guided specialist teams.
- [v2026.9.6](https://docs.openclaw.ai/releases/2026.9.6): Managed updates, restart recovery, complete 30-day Usage reporting, GitHub and remote workspace tools, live meeting notes, and new model choices.
- [v2026.9.7](https://docs.openclaw.ai/releases/2026.9.7): Snappier under load, smoother long conversations, update recovery, OpenAI Agents API, and Sign in with ChatGPT (Beta).
- [v2026.9.8](https://docs.openclaw.ai/releases/2026.9.8): Update recovery, Windows startup and reply fixes.
- [v2026.9.9](https://docs.openclaw.ai/releases/2026.9.9): GPT-6.1 Sol in Codex, Claude Haiku 5.5 support, update recovery, and fixes for replies and scheduled jobs.
- [Contributing to the threat model](https://docs.openclaw.ai/security/CONTRIBUTING-THREAT-MODEL): How to contribute to the OpenClaw threat model
- [Formal verification (security models)](https://docs.openclaw.ai/security/formal-verification): Machine-checked security models for OpenClaw's highest-risk paths.
- [Incident response](https://docs.openclaw.ai/security/incident-response): How OpenClaw triages, responds to, and follows up on security incidents
- [Network proxy](https://docs.openclaw.ai/security/network-proxy): How to route OpenClaw runtime HTTP and WebSocket traffic through an operator-managed filtering proxy
- [Threat model (MITRE ATLAS)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS): OpenClaw threat model mapped to the MITRE ATLAS framework
- [Collection and exfiltration (AML.TA0009, AML.TA0010)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/collection-and-exfiltration): OpenClaw collection and exfiltration threats (AML.TA0009, AML.TA0010): T-EXFIL-001, T-EXFIL-002, T-EXFIL-003
- [Defense evasion (AML.TA0007)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/defense-evasion): OpenClaw defense evasion threats (AML.TA0007): T-EVADE-001, T-EVADE-002
- [Discovery (AML.TA0008)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/discovery): OpenClaw discovery threats (AML.TA0008): T-DISC-001, T-DISC-002
- [Execution (AML.TA0005)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/execution): OpenClaw execution threats (AML.TA0005): T-EXEC-001, T-EXEC-002, T-EXEC-003, T-EXEC-004
- [Impact (AML.TA0011)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/impact): OpenClaw impact threats (AML.TA0011): T-IMPACT-001, T-IMPACT-002, T-IMPACT-003
- [Initial access (AML.TA0004)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/initial-access): OpenClaw initial access threats (AML.TA0004): T-ACCESS-001, T-ACCESS-002, T-ACCESS-003
- [Persistence (AML.TA0006)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/persistence): OpenClaw persistence threats (AML.TA0006): T-PERSIST-001, T-PERSIST-002, T-PERSIST-003
- [Reconnaissance (AML.TA0002)](https://docs.openclaw.ai/security/THREAT-MODEL-ATLAS/reconnaissance): OpenClaw reconnaissance threats (AML.TA0002): T-RECON-001, T-RECON-002
- [Codex supervision](https://docs.openclaw.ai/specs/codex-supervision): Architecture and product boundary for supervising native Codex sessions from OpenClaw.
- [Agent bootstrapping](https://docs.openclaw.ai/start/bootstrapping): Agent bootstrapping ritual that seeds the workspace and identity files
- [Docs directory](https://docs.openclaw.ai/start/docs-directory): Curated links to the most used OpenClaw docs.
- [Getting started](https://docs.openclaw.ai/start/getting-started): Get OpenClaw installed and run your first chat in minutes.
- [Docs hubs](https://docs.openclaw.ai/start/hubs): Hubs that link the main OpenClaw docs
- [OpenClaw lore](https://docs.openclaw.ai/start/lore): Backstory and lore of OpenClaw for context and tone
- [Onboarding (macOS app)](https://docs.openclaw.ai/start/onboarding): First-run setup flow for OpenClaw (macOS app)
- [Onboarding overview](https://docs.openclaw.ai/start/onboarding-overview): Overview of OpenClaw onboarding options and flows
- [Personal assistant setup](https://docs.openclaw.ai/start/openclaw): End-to-end guide for running OpenClaw as a personal assistant with safety cautions
- [Setup](https://docs.openclaw.ai/start/setup): Advanced setup and development workflows for OpenClaw
- [Team setup](https://docs.openclaw.ai/start/teams): Set up a shared OpenClaw gateway for a team: workspace chat, shared sessions, ownership, and roles
- [Why OpenClaw](https://docs.openclaw.ai/start/why-openclaw): The architecture case for OpenClaw: a trusted gateway, untrusted execution, deterministic policy, and versioned state, compared source-by-source with single-process harnesses
- [Identity and roles](https://docs.openclaw.ai/start/why-openclaw/identity-and-roles): Pairing, operator scopes, role ceilings, and the one-gateway-is-one-trust-domain boundary
- [OpenClaw and Hermes Agent](https://docs.openclaw.ai/start/why-openclaw/openclaw-and-hermes-agent): Source-by-source comparison of OpenClaw and Hermes Agent at the reviewed August 2026 snapshots, with the property table
- [Policy as code](https://docs.openclaw.ai/start/why-openclaw/policy-as-code): Permission modes, tool policy, and exec approvals as deterministic code-enforced gates rather than prompt requests
- [Provenance](https://docs.openclaw.ai/start/why-openclaw/provenance): Memory origin classes, trust gating, deletion boundaries, and the audit ledger's recorded non-claims
- [Secrets](https://docs.openclaw.ai/start/why-openclaw/secrets): SecretRefs, the protected store, egress sentinels, and the custody limits of each
- [The trust boundary](https://docs.openclaw.ai/start/why-openclaw/the-trust-boundary): Where OpenClaw's Gateway authority ends and configured sandbox, node, and cloud-worker execution begins
- [Versioned state, guarded upgrades](https://docs.openclaw.ai/start/why-openclaw/versioned-state-guarded-upgrades): Database-first state, schema version contracts, guarded updates, release channels, and the maturity scorecard
- [Onboarding (CLI)](https://docs.openclaw.ai/start/wizard): CLI onboarding: quick start with detected AI access or choose custom setup
- [CLI automation](https://docs.openclaw.ai/start/wizard-cli-automation): Scripted onboarding and agent setup for the OpenClaw CLI
- [CLI setup reference](https://docs.openclaw.ai/start/wizard-cli-reference): Step-by-step behavior for openclaw onboard: what each step does, config it writes, and internals
- [Overview](https://docs.openclaw.ai/tools): OpenClaw tools, skills, and plugins overview: what agents can call and how to extend them
- [ACP agents](https://docs.openclaw.ai/tools/acp-agents): Index of the OpenClaw ACP agents documentation, one page per reader job
- [ACP agents — setup](https://docs.openclaw.ai/tools/acp-agents-setup): Setting up ACP agents: acpx harness config, plugin setup, permissions
- [ACP agents bindings](https://docs.openclaw.ai/tools/acp-agents/bindings): Bind a conversation or thread to an ACP session, and configure persistent bindings[] entries
- [ACP agents controls](https://docs.openclaw.ai/tools/acp-agents/controls): Session target resolution and the full /acp command and runtime option reference
- [ACP agents delivery model](https://docs.openclaw.ai/tools/acp-agents/delivery): How ACP output is delivered for interactive and parent-owned sessions, and the sandbox boundary
- [ACP agents quickstart](https://docs.openclaw.ai/tools/acp-agents/quickstart): Install the acpx ACP runtime plugin, confirm it is usable, and pick a harness target id
- [ACP agents operator runbook](https://docs.openclaw.ai/tools/acp-agents/runbook): Day-to-day /acp flow from chat, ACP session lifecycle, and native Codex versus ACP routing
- [ACP agents sessions](https://docs.openclaw.ai/tools/acp-agents/sessions): Start ACP sessions from sessions_spawn or /acp spawn, with the parameter and mode reference
- [ACP agents troubleshooting](https://docs.openclaw.ai/tools/acp-agents/troubleshooting): Symptom, likely cause, and fix for ACP backend, plugin wiring, and delivery failures
- [Agent send](https://docs.openclaw.ai/tools/agent-send): Run agent turns from the CLI and optionally deliver replies to channels
- [apply_patch tool](https://docs.openclaw.ai/tools/apply-patch): Apply multi-file patches with the apply_patch tool
- [Ask user](https://docs.openclaw.ai/tools/ask-user): How ask_user pauses an agent turn for a structured human decision
- [Brave search](https://docs.openclaw.ai/tools/brave-search): Brave Search API setup for web_search
- [Browser (OpenClaw-managed)](https://docs.openclaw.ai/tools/browser): Index of the OpenClaw browser documentation, one page per reader job
- [Browser control API](https://docs.openclaw.ai/tools/browser-control): OpenClaw browser control API, CLI reference, and scripting actions
- [Browser troubleshooting](https://docs.openclaw.ai/tools/browser-linux-troubleshooting): Fix Chrome/Brave/Edge/Chromium CDP startup issues for OpenClaw browser control on Linux
- [Browser login](https://docs.openclaw.ai/tools/browser-login): Manual logins for browser automation + X/Twitter posting
- [WSL2 + Windows + remote Chrome CDP troubleshooting](https://docs.openclaw.ai/tools/browser-wsl2-windows-remote-cdp-troubleshooting): Troubleshoot WSL2 Gateway + Windows Chrome remote CDP in layers
- [Browser agent tools](https://docs.openclaw.ai/tools/browser/agent-tools): The single browser tool, its actions, and the arguments an agent passes
- [Browser configuration](https://docs.openclaw.ai/tools/browser/configuration): The browser config block, tab cleanup, screenshot vision, ports, SSRF policy, and picking a Chromium binary
- [Multi-profile and existing-session attach](https://docs.openclaw.ai/tools/browser/existing-session): Named browser profiles and attaching to a running Chrome session through Chrome DevTools MCP
- [Isolation and browser selection](https://docs.openclaw.ai/tools/browser/isolation): What the managed profile isolates, which browser binary is picked, and where the control API reference lives
- [Lightweight browsers](https://docs.openclaw.ai/tools/browser/lightweight): Use an externally managed Lightpanda browser for JavaScript and DOM tasks
- [Browser profiles](https://docs.openclaw.ai/tools/browser/profiles): The built-in openclaw, user, and chrome profiles and the Control UI Browser panel
- [Remote and hosted browsers](https://docs.openclaw.ai/tools/browser/remote): Local versus remote control, the node browser proxy, and hosted CDP providers
- [Browser security](https://docs.openclaw.ai/tools/browser/security): Loopback auth for the browser control API and remote CDP credential handling
- [Browser setup](https://docs.openclaw.ai/tools/browser/setup): Quick start commands, plugin enablement, agent tool policy, and the missing-command fix
- [Browser troubleshooting](https://docs.openclaw.ai/tools/browser/troubleshooting): Separating CDP startup failures from navigation SSRF blocks, plus the platform-specific pages
- [BTW side questions](https://docs.openclaw.ai/tools/btw): Ephemeral side questions with /btw
- [Chrome Extension](https://docs.openclaw.ai/tools/chrome-extension): Chrome extension: securely automate signed-in tabs with automatic local pairing
- [Code execution](https://docs.openclaw.ai/tools/code-execution): code_execution: run sandboxed remote Python analysis with xAI
- [Code Mode](https://docs.openclaw.ai/tools/code-mode): Index of the OpenClaw Code Mode documentation, one page per reader job
- [Code Mode configuration](https://docs.openclaw.ai/tools/code-mode/configuration): Code Mode configuration fields, automatic per-model activation, and the activation order
- [Code Mode executors](https://docs.openclaw.ai/tools/code-mode/executors): Choose Node or QuickJS for Code Mode and understand their execution and security boundaries
- [Code Mode guest API](https://docs.openclaw.ai/tools/code-mode/guest-api): Globals, catalog handles, MCP namespaces, and virtual API declarations inside guest code
- [Code Mode internals](https://docs.openclaw.ai/tools/code-mode/internals): Code Mode scope, nested tool execution, executor continuations, and security boundaries
- [Code Mode maintainer notes](https://docs.openclaw.ai/tools/code-mode/maintainers): Code Mode implementation layout, the validation checklist, and the E2E test plan
- [Code Mode output](https://docs.openclaw.ai/tools/code-mode/output): Declared output contracts for tool results and the guest output API
- [Code Mode quickstart](https://docs.openclaw.ai/tools/code-mode/quickstart): Enable OpenClaw Code Mode, override one model, and recover from tool errors
- [Code Mode tool surface](https://docs.openclaw.ai/tools/code-mode/tool-surface): The model-visible exec and wait contracts, the hidden tool catalog, and name collisions
- [Code Mode troubleshooting](https://docs.openclaw.ai/tools/code-mode/troubleshooting): Code Mode error codes, telemetry fields, and the debug environment variables
- [Creating skills](https://docs.openclaw.ai/tools/creating-skills): Build, test, and publish custom SKILL.md workspace skills or personal skills on a shared Gateway.
- [Custodian skills](https://docs.openclaw.ai/tools/custodian-skills): Release-versioned operational skills that only the configured Custodian agent can discover and use.
- [Diffs](https://docs.openclaw.ai/tools/diffs): Read-only diff viewer and file renderer for agents (optional plugin tool)
- [DuckDuckGo search](https://docs.openclaw.ai/tools/duckduckgo-search): DuckDuckGo web search -- key-free provider (experimental, HTML-based)
- [Elevated mode](https://docs.openclaw.ai/tools/elevated): Elevated exec mode: run commands outside the sandbox from a sandboxed agent
- [Exa search](https://docs.openclaw.ai/tools/exa-search): Exa AI search -- neural and keyword search with content extraction
- [Exec tool](https://docs.openclaw.ai/tools/exec): Exec tool usage, stdin modes, and TTY support
- [Exec approvals](https://docs.openclaw.ai/tools/exec-approvals): Host exec approvals: policy knobs, allowlists, and the YOLO/strict workflow
- [Exec approvals — advanced](https://docs.openclaw.ai/tools/exec-approvals-advanced): Advanced exec approvals: safe bins, interpreter binding, approval forwarding, native delivery
- [Firecrawl](https://docs.openclaw.ai/tools/firecrawl): Firecrawl search, scrape, and web_fetch fallback
- [Gemini search](https://docs.openclaw.ai/tools/gemini-search): Gemini web search with Google Search grounding
- [Goal](https://docs.openclaw.ai/tools/goal): Session goals: durable per-session objectives, /goal controls, model goal tools, token budgets, and TUI status
- [Grok search](https://docs.openclaw.ai/tools/grok-search): Grok web search via xAI web-grounded responses
- [Image generation](https://docs.openclaw.ai/tools/image-generation): Generate and edit images via image_generate across OpenAI, Google, fal, Microsoft Foundry, MiniMax, ComfyUI, DeepInfra, OpenRouter, LiteLLM, xAI, Vydra
- [Kimi search](https://docs.openclaw.ai/tools/kimi-search): Kimi web search via Moonshot web search
- [LLM task](https://docs.openclaw.ai/tools/llm-task): JSON-only LLM tasks for workflows (optional plugin tool)
- [Lobster](https://docs.openclaw.ai/tools/lobster): Typed workflow runtime for OpenClaw with resumable approval and input gates.
- [Tool-loop detection](https://docs.openclaw.ai/tools/loop-detection): How to enable guardrails that detect repetitive tool-call loops
- [Connect MCP servers](https://docs.openclaw.ai/tools/mcp): Connect MCP servers to OpenClaw from the Control UI, CLI, or config
- [Media overview](https://docs.openclaw.ai/tools/media-overview): Image, video, music, speech, and media-understanding capabilities at a glance
- [MiniMax search](https://docs.openclaw.ai/tools/minimax-search): MiniMax Search via the Token Plan search API
- [Multi-agent sandbox and tools](https://docs.openclaw.ai/tools/multi-agent-sandbox-tools): Per-agent sandbox + tool restrictions, precedence, and examples
- [Music generation](https://docs.openclaw.ai/tools/music-generation): Generate music via music_generate across ComfyUI, fal, Google Lyria, MiniMax, and OpenRouter workflows
- [Ollama web search](https://docs.openclaw.ai/tools/ollama-search): Ollama Web Search via a local Ollama host or the hosted Ollama API
- [Parallel search](https://docs.openclaw.ai/tools/parallel-search): Parallel Search -- LLM-optimized dense excerpts from web sources
- [PDF tool](https://docs.openclaw.ai/tools/pdf): Analyze one or more PDF documents with native provider support and extraction fallback
- [Permission modes](https://docs.openclaw.ai/tools/permission-modes): Permission modes for host exec, Codex Guardian approvals, and ACPX harness sessions
- [Perplexity search](https://docs.openclaw.ai/tools/perplexity-search): Perplexity Search API and Sonar/OpenRouter compatibility for web_search
- [Plugins](https://docs.openclaw.ai/tools/plugin): Install, configure, and manage OpenClaw plugins
- [Progress card](https://docs.openclaw.ai/tools/progress-card): Maintain one durable plan and status card for a session
- [Reactions](https://docs.openclaw.ai/tools/reactions): Reaction tool semantics across all supported channels
- [Screen](https://docs.openclaw.ai/tools/screen): Let an agent arrange the connected Control UI
- [SearXNG search](https://docs.openclaw.ai/tools/searxng-search): SearXNG web search -- self-hosted, key-free meta-search provider
- [Secrets](https://docs.openclaw.ai/tools/secrets): How the secrets tool lets the agent request credentials it never sees
- [Self-learning](https://docs.openclaw.ai/tools/self-learning): How OpenClaw reviews finished work in the background and saves lessons as learned skills
- [Show widget](https://docs.openclaw.ai/tools/show-widget): Show HTML widgets on supported chat surfaces or pin native data reports to dashboards
- [Skill Workshop](https://docs.openclaw.ai/tools/skill-workshop): Skills your agent learns on its own: how they are saved, undo, unused-skill cleanup, /learn, config, storage, and operator surfaces
- [Personal library authoring](https://docs.openclaw.ai/tools/skill-workshop/personal-library): Create and update profile-owned personal skills on a shared Gateway
- [Skills](https://docs.openclaw.ai/tools/skills): Skills teach your agent how to use tools. Learn how they load, how precedence works, and how to configure gating, allowlists, and environment injection.
- [Skills config](https://docs.openclaw.ai/tools/skills-config): Full reference for the skills.* config schema, agent allowlists, workshop settings, and sandbox env var handling.
- [Slash commands](https://docs.openclaw.ai/tools/slash-commands): All available slash commands, directives, and inline shortcuts — configuration, routing, and per-surface behavior.
- [Steer](https://docs.openclaw.ai/tools/steer): Steer an active run without changing queue mode
- [Sub-agents](https://docs.openclaw.ai/tools/subagents): Index of the OpenClaw sub-agent documentation, one page per reader job
- [Sub-agent announce](https://docs.openclaw.ai/tools/subagents/announce): How a sub-agent reports its result back, the announce context block, and why sessions_history is preferred
- [Nested sub-agents and authentication](https://docs.openclaw.ai/tools/subagents/nesting): Recursive delegation depth, the announce chain, cascade stop, and how sub-agent auth and memory audience resolve
- [Sub-agent concurrency, recovery, and stopping](https://docs.openclaw.ai/tools/subagents/operations): The subagent queue lane, restart recovery, stop scope, and the standing limitations
- [Sub-agent slash command](https://docs.openclaw.ai/tools/subagents/slash-command): Inspect sub-agent runs with /subagents, use the thread-binding commands, and follow the completion-delivery path
- [Thread-bound sub-agent sessions](https://docs.openclaw.ai/tools/subagents/thread-bound-sessions): Bind a sub-agent to a channel thread, and the allowlist, discovery, and auto-archive rules
- [Sub-agent tool policy](https://docs.openclaw.ai/tools/subagents/tool-policy): The sub-agent tool restriction layer and how to narrow it with config
- [Sub-agent tool reference](https://docs.openclaw.ai/tools/subagents/tool-reference): Context modes and the sessions_spawn, sessions_yield, and subagents tool contracts
- [Swarm](https://docs.openclaw.ai/tools/swarm): Orchestrate concurrent sub-agents from Code Mode scripts with structured results, bounded fan-out, and live progress
- [Tavily](https://docs.openclaw.ai/tools/tavily): Tavily search and extract tools
- [Theme](https://docs.openclaw.ai/tools/theme): Let an agent select plugin themes or create a personal OpenClaw theme
- [Thinking levels](https://docs.openclaw.ai/tools/thinking): Directive syntax for /think, /fast, /verbose, /trace, and reasoning visibility
- [Tokenjuice](https://docs.openclaw.ai/tools/tokenjuice): Compact noisy exec and bash tool results with the optional Tokenjuice plugin
- [Tool Search](https://docs.openclaw.ai/tools/tool-search): Tool Search: compact large OpenClaw tool catalogs behind search, describe, and call
- [Trajectory bundles](https://docs.openclaw.ai/tools/trajectory): Export redacted trajectory bundles for debugging an OpenClaw agent session
- [Text-to-speech](https://docs.openclaw.ai/tools/tts): Index of the OpenClaw text-to-speech documentation, one page per reader job
- [Text-to-speech agent tool and Gateway RPC](https://docs.openclaw.ai/tools/tts/api): The tts agent tool and the Gateway TTS RPC methods
- [Text-to-speech commands and directives](https://docs.openclaw.ai/tools/tts/commands): Model-emitted TTS directives, the /tts slash commands, and per-user preferences
- [Text-to-speech configuration](https://docs.openclaw.ai/tools/tts/configuration): The tts config block, per-provider settings, local speech engines, and override precedence
- [Text-to-speech field reference](https://docs.openclaw.ai/tools/tts/field-reference): Every tts.* and tts.providers.<id>.* configuration field
- [Text-to-speech output and Auto-TTS behavior](https://docs.openclaw.ai/tools/tts/output): Per-channel audio formats, transcoding, and what Auto-TTS sends
- [Text-to-speech personas](https://docs.openclaw.ai/tools/tts/personas): Persona definitions, provider bindings, resolution order, and fallback policy
- [Text-to-speech quickstart](https://docs.openclaw.ai/tools/tts/quickstart): Turn on text-to-speech, pick a provider, and send a first audio reply
- [Video generation](https://docs.openclaw.ai/tools/video-generation): Generate videos via video_generate from text, image, or video references across 18 provider backends
- [Web search](https://docs.openclaw.ai/tools/web): web_search, x_search, and web_fetch -- search the web, search X posts, or fetch page content
- [Web fetch](https://docs.openclaw.ai/tools/web-fetch): web_fetch tool -- HTTP fetch with readable content extraction
- [Linux server](https://docs.openclaw.ai/vps): Run OpenClaw on a Linux server or cloud VPS — provider picker, architecture, and tuning
- [Web](https://docs.openclaw.ai/web): Gateway web surfaces: Control UI, bind modes, and security
- [Control UI](https://docs.openclaw.ai/web/control-ui): Browser-based control UI for the Gateway (chat, activity, nodes, config)
- [Chat](https://docs.openclaw.ai/web/control-ui/chat): Composer controls, transcript rendering, side chat, and hosted embeds
- [Connect and pair](https://docs.openclaw.ai/web/control-ui/connect-and-pair): Pair browsers and phones with the Gateway and reach the Control UI over Tailscale
- [Build and develop](https://docs.openclaw.ai/web/control-ui/development): Build the Control UI, run the dev server, and point it at a remote Gateway
- [Feature and RPC reference](https://docs.openclaw.ai/web/control-ui/feature-reference): What the Control UI can do today, grouped by capability with its Gateway RPC names
- [Offline and reconnect](https://docs.openclaw.ai/web/control-ui/offline-and-reconnect): What the Control UI keeps when the Gateway connection drops, and how it recovers
- [Panels and docks](https://docs.openclaw.ai/web/control-ui/panels): Ask OpenClaw, the Home dock, the operator terminal, and the browser panel
- [Security model](https://docs.openclaw.ai/web/control-ui/security-model): Content security policy, public transcript boundaries, authenticated media routes, and approval links
- [Sessions and sidebar](https://docs.openclaw.ai/web/control-ui/sessions-and-sidebar): Sidebar zones, session menus, and the New session page
- [Settings](https://docs.openclaw.ai/web/control-ui/settings): Identity, appearance, plugins, updates, MCP, activity, and meetings
- [Dashboard](https://docs.openclaw.ai/web/dashboard): Gateway dashboard (Control UI) access and auth
- [Dashboard Architecture](https://docs.openclaw.ai/web/dashboard-architecture): Session dashboard architecture: widget hosting, capabilities, board storage, and protocol
- [Session Dashboards](https://docs.openclaw.ai/web/dashboards): Session dashboards: agent-built widgets, boards, tabs, and flexible task layouts
- [The Lobster](https://docs.openclaw.ai/web/lobster): A lobster sometimes visits the Control UI. This page explains just enough.
- [Notifications](https://docs.openclaw.ai/web/notifications): Enable and test browser or macOS notifications from the Control UI
- [TUI](https://docs.openclaw.ai/web/tui): Terminal UI (TUI): connect to the Gateway or run locally in embedded mode
- [Control UI URLs](https://docs.openclaw.ai/web/urls): Control UI routes, focus presentations, stable session links, and connection handoff parameters
- [WebChat](https://docs.openclaw.ai/web/webchat): Native and Control UI WebChat usage over the Gateway WebSocket