# DevTune - [Introduction to DevTune](https://docs.devtune.ai/introduction.md): See what AI platforms say about your product and who they cite, what AI systems fetch from your site, which fixes to make, and what moved after you shipped them. - [Quickstart](https://docs.devtune.ai/getting-started/quickstart.md): Sign up, create your first project with the activation wizard, start your trial, and read your first Visibility results. - [Set up DevTune with an AI agent](https://docs.devtune.ai/getting-started/agent-setup.md): Copy one prompt into Claude Code, Codex, Cursor or another AI agent. It sets up your DevTune project, and you confirm it with a 6-digit code. - [Create your account](https://docs.devtune.ai/getting-started/creating-account.md): Sign up with email, Google or GitHub, choose a plan, and set up your team workspace. - [Key concepts](https://docs.devtune.ai/getting-started/key-concepts.md): Look up the core DevTune terms: projects, prompts, brands, tracked URLs, citations, share of voice, presence rate and Visibility metrics. - [Account and project layout](https://docs.devtune.ai/getting-started/dashboard-overview.md): Get oriented in DevTune: the account project list, Project Home with coverage, Get started, headline tiles and cards, and the project sidebar. - [Troubleshooting](https://docs.devtune.ai/troubleshooting.md): Find the cause of a common DevTune problem by its symptom, and go to the page with the steps that fix it. - [Use DevTune with your agent](https://docs.devtune.ai/your-agent/overview.md): Connect Claude Code, Codex, Muse Code, Cursor and other agents to your DevTune project, then measure, fix and prove from inside the agent. - [Use DevTune with Claude Code](https://docs.devtune.ai/your-agent/claude-code.md): Connect Claude Code to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Claude Code. - [Use DevTune with Codex](https://docs.devtune.ai/your-agent/codex.md): Connect Codex to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Codex. - [Use DevTune with Muse Code](https://docs.devtune.ai/your-agent/muse-code.md): Connect Muse Code to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Muse Code. - [Use DevTune with Cursor](https://docs.devtune.ai/your-agent/cursor.md): Connect Cursor to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Cursor. - [Use DevTune with Gemini CLI](https://docs.devtune.ai/your-agent/gemini-cli.md): Connect Gemini CLI to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Gemini CLI. - [Use DevTune with OpenCode](https://docs.devtune.ai/your-agent/opencode.md): Connect OpenCode to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside OpenCode. - [Use DevTune with Pi](https://docs.devtune.ai/your-agent/pi.md): Connect Pi to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Pi. - [Use DevTune with Droid](https://docs.devtune.ai/your-agent/droid.md): Connect Droid to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Droid. - [Use DevTune with Devin CLI](https://docs.devtune.ai/your-agent/devin.md): Connect Devin CLI to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Devin CLI. - [Use DevTune with Claude](https://docs.devtune.ai/your-agent/claude.md): Connect Claude to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Claude. - [Use DevTune with ChatGPT](https://docs.devtune.ai/your-agent/chatgpt.md): Connect ChatGPT to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside ChatGPT. - [Use DevTune with Hermes Agent](https://docs.devtune.ai/your-agent/hermes.md): Connect Hermes Agent to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Hermes Agent. - [Use DevTune with Eve](https://docs.devtune.ai/your-agent/eve.md): Connect Eve to your DevTune project, then measure what AI says about your product, pick the next fix and check the result from inside Eve. - [Overview](https://docs.devtune.ai/agent-picks/overview.md): Measure which tool coding agents choose when asked to add a capability to an app. - [Set up](https://docs.devtune.ai/agent-picks/set-up.md): Choose your category or job, tools, test apps, requests and models. - [Starter test](https://docs.devtune.ai/agent-picks/starter-test.md): Run your free first test and read the individual picks as they finish. - [Paid tests and credits](https://docs.devtune.ai/agent-picks/paid-tests-and-credits.md): Compare Agent Picks plans, review estimates, reserve credits and check the final charge for each test. - [Schedules](https://docs.devtune.ai/agent-picks/schedules.md): Schedule paid tests, confirm a credit limit and recover from skipped runs. - [Reading results](https://docs.devtune.ai/agent-picks/reading-results.md): Read pick rates, install rates, likely ranges, individual runs and incomplete coverage. - [Measurement methodology](https://docs.devtune.ai/agent-picks/methodology.md): Learn what an Agent Picks test measures, how it classifies outcomes, and how accurate and precise the classifier is in each category. - [Losses, Actions and re-tests](https://docs.devtune.ai/agent-picks/losses-actions-and-retests.md): Review losing scenarios, draft a fix and measure installs after shipping it. - [Run tests on your machine](https://docs.devtune.ai/agent-picks/run-on-your-machine.md): Write a setup file with the agentpicks CLI, test the coding agents installed on your machine, read which tool each run picked, and upload the runs to your DevTune project. - [Run tests in GitHub Actions](https://docs.devtune.ai/agent-picks/github-action.md): Run an Agent Picks setup on a schedule in GitHub Actions and upload each run to your DevTune project. - [API and MCP](https://docs.devtune.ai/agent-picks/api-and-mcp.md): Read Agent Picks measurements, run tests and manage the setup and schedules from your own tools. - [Actions and Results](https://docs.devtune.ai/insights-actions/overview.md): Learn how DevTune turns Visibility, Audit and AI Traffic signals into a ranked queue of fixes, and measures what moved after each one went live. - [Actions queue](https://docs.devtune.ai/insights-actions/actions-hub.md): Work the ranked Actions queue: accept suggestions, generate briefs, mark fixes live, and follow each action through to its measured result. - [Content drafts](https://docs.devtune.ai/insights-actions/content-drafts.md): Generate a page draft or section edits from an action brief, check their facts, and download the markdown to publish yourself. - [Agents](https://docs.devtune.ai/agents/overview.md): Run project-scoped agents, review their draft outputs, schedule repeat work, and follow run history. - [Custom agents](https://docs.devtune.ai/agents/custom-agents.md): Create a project agent with your own instructions, safe data tools, reviewable output, model, and schedule. - [Agent models and credits](https://docs.devtune.ai/agents/models-and-credits.md): Choose an agent model, read multiplier bands and run estimates, and understand credit reservation and settlement. - [Library](https://docs.devtune.ai/library/overview.md): Use the project Library to find agent deliverables, uploaded source material, version history, and saved memory. - [Library uploads](https://docs.devtune.ai/library/uploads.md): Upload supported documents, replace an upload with a new version, and interpret extraction and indexing states. - [Library search and agent grounding](https://docs.devtune.ai/library/search-and-grounding.md): Find Library items with keyword and semantic search, filter the catalog, and control which items agents can retrieve. - [Project context](https://docs.devtune.ai/library/knowledge.md): Manage the company context and agent preferences shared across your project. - [Results](https://docs.devtune.ai/insights-actions/outcomes.md): Read what moved after an action went live: measured actions, before-and-after page signals, and content changes no action claims. - [Shared reports](https://docs.devtune.ai/projects/shared-reports.md): Send a read-only report of a project's results to someone without a DevTune account using an expiring share link, and revoke it when it is no longer needed. - [Alerts and notifications](https://docs.devtune.ai/insights-actions/alerts.md): Configure DevTune project alerts for visibility regressions, competitor threats, and weekly digests, and review them under Notifications. - [Visibility overview](https://docs.devtune.ai/search-tracking/overview.md): Monitor how often AI platforms mention and cite your brand, your share of voice, and who they cite instead. - [Set up Visibility](https://docs.devtune.ai/search-tracking/getting-started.md): Configure brands, tracked URLs, brand terms, prompts and platforms so DevTune can start measuring what AI platforms say about your project. - [Prompts](https://docs.devtune.ai/search-tracking/prompts.md): Manage the prompts DevTune submits to AI platforms each run and inspect prompt-level responses, citations, brand mentions, and competitive context. - [Topics](https://docs.devtune.ai/search-tracking/topics.md): Group prompts into topics to analyse visibility by theme and to weight action recommendations and opportunity scores. - [Prompt-set analysis](https://docs.devtune.ai/search-tracking/prompt-set-analysis.md): Review your tracked prompts and approve changes individually. - [Tracked URLs](https://docs.devtune.ai/search-tracking/domains.md): Configure the tracked URLs and domains DevTune checks for citations and links in AI answers, managed per brand from Settings. - [Brand Terms](https://docs.devtune.ai/search-tracking/keywords.md): Manage the brand-name terms DevTune uses to detect your product and competitor mentions in AI answers, configured per brand. - [Writing positioning statements](https://docs.devtune.ai/search-tracking/writing-statements.md): Write positioning statements that AI answers can actually be measured against: one short quality, in your buyers’ words, with the right prompt intents. - [Visibility dashboard](https://docs.devtune.ai/search-tracking/search-overview-dashboard.md): Read the Visibility Dashboard: Presence Rate with its range and lead line, Average Position, Sentiment, Positioning, advertising, the trend chart, Market comparison, top domains, and strengths and gaps. - [Citations](https://docs.devtune.ai/search-tracking/citations.md): See which URLs AI platforms cite, whether each one is yours or a third party, how domains rank, and the detail behind each citation. - [Competitors](https://docs.devtune.ai/search-tracking/competitors.md): See which competing brands AI platforms cite, the prompts where they are cited and you are absent, and the topics they lead on. - [Positioning](https://docs.devtune.ai/search-tracking/positioning.md): Measure which product AI answers credit with each of your positioning statements, how that share moves over time, and the answers behind every number. - [Accuracy](https://docs.devtune.ai/search-tracking/accuracy.md): Check what AI answers say about your product against your own site, see which claims contradict it, and fix them. - [Lookalike sites](https://docs.devtune.ai/search-tracking/lookalike-sites.md): Find the sites AI answers cite in place of yours, with domains that imitate your own, and report them to their registrar and the AI platforms. - [Platform comparison](https://docs.devtune.ai/search-tracking/platform-comparison.md): Compare visibility across AI platforms to find platform-specific gaps and decide where to invest in content. - [Explore](https://docs.devtune.ai/analytics/search-analytics.md): Chart how often AI answers cite your docs, blog, GitHub and other properties over time, and which types of page and source verticals AI platforms cite. - [Visibility metrics](https://docs.devtune.ai/search-tracking/visibility-metrics.md): Learn how DevTune measures visibility: presence rate, share of voice, sentiment, average position, citation metrics, and the 90% ranges and lead rule behind them. - [How Visibility runs work](https://docs.devtune.ai/search-tracking/running-search-tests.md): Learn how DevTune schedules Visibility runs, what each run measures, and how to read presence, citation and brand mention results. - [Audit](https://docs.devtune.ai/audit/overview.md): Use domain readiness, page evidence, diagnoses, and findings to decide what needs attention. - [AI Traffic overview](https://docs.devtune.ai/ai-traffic/overview.md): See what AI systems crawl and fetch from your site, and the AI referral sessions and conversions you can attribute. - [Choose your sensors](https://docs.devtune.ai/ai-traffic/choosing-sensors.md): Compare what each sensor measures, choose a crawler sensor, and see how probes, the snippet, Search Console and website analytics complete coverage. - [Set up AI Traffic](https://docs.devtune.ai/ai-traffic/setup.md): Connect the sensors that feed AI Traffic: Cloudflare pull, edge middleware or a Vercel log drain, GA4 or PostHog, Search Console, and the tracking snippet. - [Connect Cloudflare pull](https://docs.devtune.ai/ai-traffic/cloudflare.md): Measure AI crawls, answer fetches and 404 demand on a Cloudflare-proxied hostname with a read-only API token. No code changes. - [Install edge middleware](https://docs.devtune.ai/ai-traffic/edge-middleware.md): Record AI crawls, answer fetches, 404 demand and AI referral visits (version 0.13.0 or later) from your Next.js, Express or Node app with @devtune/ai-traffic, using an ingest key or a linked Vercel project. - [Connect a Vercel log drain](https://docs.devtune.ai/ai-traffic/vercel-log-drain.md): Measure AI crawls, answer fetches and 404 demand on your Vercel projects with a log drain. No code changes. - [Install the tracking snippet](https://docs.devtune.ai/ai-traffic/tracking-snippet.md): Add the DevTune tracking snippet to your site for a first-party record of AI referral page views, and proxy it through your own domain if needed. - [Advanced sensor settings](https://docs.devtune.ai/ai-traffic/advanced-setup.md): Configure crawler identity checks, trusted proxies, delivery, response statuses and package versions for the AI Traffic crawler sensors. - [Connect PostHog](https://docs.devtune.ai/ai-traffic/posthog.md): Import AI referral sessions and selected conversions from PostHog Cloud. - [Reading AI Traffic](https://docs.devtune.ai/ai-traffic/dashboard.md): Read the AI Traffic Overview, Crawls & fetches and Referrals views, and interpret crawls, answer fetches, AI referrals and conversions. - [Detected AI Traffic changes](https://docs.devtune.ai/ai-traffic/detected-changes.md): Learn when DevTune reports that an AI platform started or stopped fetching or crawling your site, the thresholds it uses, and why a sensor outage never shows as a drop. - [Direct traffic](https://docs.devtune.ai/ai-traffic/direct-traffic.md): Read the GA4 direct-traffic chart and understand when DevTune marks a change as AI-correlated lift. - [Working with projects](https://docs.devtune.ai/projects/overview.md): Organize and manage DevTune projects: each project runs the Visibility, AI Traffic, Actions and Results loop for one product. - [Navigating your project](https://docs.devtune.ai/projects/project-navigation.md): Map of the DevTune project sidebar: Home, Actions, Agents, Results, the Insights pages for Visibility, Audit and Traffic, and Settings. - [Settings](https://docs.devtune.ai/projects/project-settings.md): Manage your project's general setup, company context, brands, platforms, content preferences, locked pages, report sharing and sensor integrations in Settings. - [Project members and roles](https://docs.devtune.ai/projects/project-members.md): Manage DevTune project access through team membership, including how shared team-account access works, role permissions, and inviting new members. - [Account and billing](https://docs.devtune.ai/account-billing/overview.md): Manage team membership, plans, shared credits, API keys, webhooks and API activity for your account. - [Credit transactions](https://docs.devtune.ai/account-billing/credit-transactions.md): Read the account Transactions ledger for allowances, purchases, content generation, and agent-run costs. - [API Keys](https://docs.devtune.ai/account-billing/api-keys.md): Create, scope, expire, and revoke project-scoped DevTune API keys for the public REST API and MCP server, all managed from the account sidebar. - [API Activity](https://docs.devtune.ai/account-billing/agent-activity.md): Audit recent DevTune API and AI agent traffic across your account by API key, project, endpoint, scope, status code, and request latency. - [Personal settings and notifications](https://docs.devtune.ai/account-billing/account-settings.md): Set your profile, notification preferences, appearance, password and MFA, and override notifications for a single project. - [Team management](https://docs.devtune.ai/account-billing/team-management.md): Manage your DevTune team workspace settings, members, invitations, roles, domain auto-join, and ownership transfer from the Team page. - [Inactive data deletion](https://docs.devtune.ai/account-billing/inactive-data-deletion.md): When DevTune deletes paused projects, team accounts without a plan, and logins that belong to no team, how you are warned, and how to keep your data. - [Slack agent conversations](https://docs.devtune.ai/integrations/slack-agent-channels.md): Connect a Slack workspace, link member identities, chat with Dex in threads, confirm specialist agent runs, and review their drafts. - [GitHub pull requests](https://docs.devtune.ai/integrations/github-pull-requests.md): Connect GitHub, map your site to content repositories, and open pull requests from action drafts. Merging marks the action live. - [GitBook change requests](https://docs.devtune.ai/integrations/gitbook-change-requests.md): Connect GitBook with a personal access token, map your site to GitBook spaces, and open change requests from action drafts. Merging marks the action live. - [REST API reference](https://docs.devtune.ai/api-reference/overview.md): Pull Visibility, citation, AI Traffic, Actions and Results data into CI/CD pipelines, BI tools and dashboards with the public REST API. - [Prompt-set analysis API and MCP](https://docs.devtune.ai/api-reference/prompt-set-analysis.md): Read prompt assessments, start an approved analysis and review proposals through REST or MCP. - [API authentication](https://docs.devtune.ai/api-reference/authentication.md): Create project-scoped API keys, configure endpoint scopes, and authenticate DevTune REST API requests with the Authorization bearer header. - [Human approval](https://docs.devtune.ai/api-reference/human-approval.md): Which steps only a person can take in DevTune, which a credential can take on its own, and which approvals the caller attests to. - [Agent Registration](https://docs.devtune.ai/api-reference/agent-registration.md): Let an AI agent register with DevTune through auth.md, draft a project setup, and act on the project once a person claims it. - [Try the demo](https://docs.devtune.ai/api-reference/demo.md): Read real DevTune data over MCP or the REST API without an account, using the public demo token on DevTune’s category projects. - [API Rate Limits](https://docs.devtune.ai/api-reference/rate-limits.md): DevTune REST API request limits per minute by plan, the separate agent-run limit, rate-limit headers and 429 handling. - [API errors](https://docs.devtune.ai/api-reference/errors.md): Look up DevTune API status codes, the error response format and common error codes, and decide when to retry a failed request. - [Probe health](https://docs.devtune.ai/api-reference/probe-health.md): Check which AI platforms DevTune probed successfully for a project, when each last answered, and any open platform outage, before trusting or retrying a Visibility read that looks stale. - [MCP server](https://docs.devtune.ai/api-reference/mcp-server.md): Connect Codex, Claude, Cursor and other MCP clients to your DevTune project data through the Streamable HTTP Model Context Protocol server. - [Code mode operations](https://docs.devtune.ai/api-reference/mcp-code-mode.md): Scopes, approval rules, receipts and recovery for each write operation in DevTune MCP code mode. - [Webhooks](https://docs.devtune.ai/api-reference/webhooks.md): Receive signed HTTP notifications when visibility changes, a Visibility run completes, an action changes, a brief, draft or agent run finishes, a result is measured, or an AI Traffic sensor stops or recovers, with payload examples and signature verification. - [TypeScript client](https://docs.devtune.ai/sdk/typescript.md): Use the project-scoped DevTune client to read Actions, citation coverage, AI Traffic and Results, or launch an explicitly intended managed-agent task. - [CLI](https://docs.devtune.ai/sdk/cli.md): Read project measurements, Actions and briefs, launch an explicitly approved managed-agent task, and wait for an existing task. - [Discover the authorized project](https://docs.devtune.ai/api-reference/discover-the-authorized-project.md): Returns the same connection context as the project-bound route, using the project already selected during OAuth consent or bound to the API key. Never lists other projects. Never cached. Requires the `project.read` scope. - [Get the connection context](https://docs.devtune.ai/api-reference/get-the-connection-context.md): Returns the project this API key or OAuth grant is bound to, its own domains, the scopes the credential holds now and which account features are on. A paused project keeps only read scopes and setup.write, which can resume it. An operation needs its scope and, where it has one, its feature; neither… - [Revoke the calling API key](https://docs.devtune.ai/api-reference/revoke-the-calling-api-key.md): Revokes the calling API key. Every later request with it gets 401. Needs only project.read, which every API key holds, and acts on no other key. Works while the project is paused. A claimed agent's dtk_agent_ token revokes the key its claim created, so the token stops working and the agent must regi… - [Rotate the calling API key](https://docs.devtune.ai/api-reference/rotate-the-calling-api-key.md): Replaces the calling API key's secret and returns the new secret once. The key keeps its id, name, scopes, expiry and rate limit window; only the secret changes. Requests already in flight during the rotation count against the old secret. The old secret stops working in the same step: if the respons… - [Get project status and usage](https://docs.devtune.ai/api-reference/get-project-status-and-usage.md): Returns whether the project is paused and its usage: active prompts, credits spent this credit period, AI Traffic events this UTC month and tracked pages. Each figure carries its own period. Also returns when the project and its team account are scheduled to be deleted for inactivity. Every API key… - [Get code-mode operation receipts](https://docs.devtune.ai/api-reference/get-code-mode-operation-receipts.md): Returns execution status and bounded operation receipts for this same API key and project. Receipts exclude code, arguments and result data. A running or uncertain operation may have committed; never automatically repeat it. Executions are retained with the project. No other credential, including an… - [List brands](https://docs.devtune.ai/api-reference/list-brands.md): Pages the project's brands: the primary brand and the competitor brands AI Search measures against. Only tracked brands are listed unless includeArchived is true; an archived brand has archivedAt set, is no longer tracked and keeps its recorded history. Pass a tracked brand's id as brandId to create… - [Create a brand](https://docs.devtune.ai/api-reference/create-a-brand.md): Adds a primary or competitor brand to this project. A brand with the same name in any case returns 409. When that brand is archived, the 409 has reason brand_archived and details.brandId; restore it with restoreBrand instead. Concurrent brand changes return 409 with reason concurrent_brand_update an… - [Update a brand](https://docs.devtune.ai/api-reference/update-a-brand.md): Renames a brand or changes its brandType. A type change relabels its domains and brand terms in one transaction. Becoming primary is refused with 400 if a home URL is not canonical for a primary brand. Recorded citations keep their measured labels; mentions in earlier answers show the new type. Prov… - [Archive a brand](https://docs.devtune.ai/api-reference/archive-a-brand.md): Archives a competitor brand: DevTune stops tracking it from now and leaves it out of positioning comparisons. Its recorded citations, mentions, domains and brand terms stay as history. Restore it with restoreBrand. The primary brand cannot be archived and returns 409 with reason primary_brand_not_ar… - [Restore a brand](https://docs.devtune.ai/api-reference/restore-a-brand.md): Restores an archived brand: DevTune tracks it again from now, and the days it was archived stay untracked. Restoring a tracked brand succeeds again. A brand outside this project returns 404. Concurrent brand changes return 409 with reason concurrent_brand_update and retryable true; retry the request… - [List tracked domains](https://docs.devtune.ai/api-reference/list-tracked-domains.md): Pages the domains and URLs added to the project's brands, which decide whose a citation is. Domains recorded only because AI answers cited them are not listed. Ordered by id. Pass nextCursor back as cursor. This is a live listing, not a snapshot. Requires the `setup.read` scope. - [Add a domain](https://docs.devtune.ai/api-reference/add-a-domain.md): Adds a domain or URL to a brand of this project. It counts as the brand's from today; citations already recorded keep their labels. A domain the project already holds, including one recorded from citations, moves to this brand, and previousBrandId names the brand it left. A host claimed by another b… - [Update a domain](https://docs.devtune.ai/api-reference/update-a-domain.md): Changes one thing about a tracked domain per request: domain renames it using the same validation as the app, brandId moves it to another brand of this project (null unassigns it), domainType tags it, or siteType sets how Audit scores one of the primary brand's domains. Only domains added to a brand… - [Delete a domain](https://docs.devtune.ai/api-reference/delete-a-domain.md): Stops tracking a domain. Citations already recorded keep their labels. This cannot be undone; add it again with createDomain. A domain already deleted, or outside this project, returns 404. Requires the `setup.write` scope. - [List brand terms](https://docs.devtune.ai/api-reference/list-brand-terms.md): Pages the project's brand terms: the brand and product names DevTune watches for in AI answers to count mentions. Only active terms are listed unless includeInactive is true; a retired term is one deleted while assigned to a brand, kept so earlier mentions stay attributed. Ordered by id. Pass nextCu… - [Add a brand term](https://docs.devtune.ai/api-reference/add-a-brand-term.md): Adds a brand term, optionally assigned to a brand of this project. Terms are matched in any case and do not run as prompts or spend credits. Adding a retired term again under the same brand restores it and returns restored true. A term already active, or retired under another brand, returns 409. An… - [Update a brand term](https://docs.devtune.ai/api-reference/update-a-brand-term.md): Assigns a brand term to another brand of this project, or unassigns it with null. An archived brand returns 409 with reason brand_archived. A term or brand outside this project returns 404. Requires the `setup.write` scope. - [Delete a brand term](https://docs.devtune.ai/api-reference/delete-a-brand-term.md): Deletes a brand term. A term assigned to a brand is retired rather than removed, so earlier mentions stay attributed, and retired is true; adding it again restores it. An unassigned term is removed. A term already removed, or outside this project, returns 404. Requires the `setup.write` scope. - [List probed platforms](https://docs.devtune.ai/api-reference/list-probed-platforms.md): Lists the AI platforms the project can be probed on. enabled is true when scheduled probes run on the platform now: the project has it switched on, the plan includes it, and the project is not paused; Grok also needs DevTune to be collecting it. available is true when the account's plan includes the… - [List content sources](https://docs.devtune.ai/api-reference/list-content-sources.md): Lists the project's content sources, for accounts with content pull requests: each maps a site URL prefix onto a GitHub repository, branch and folder that DevTune can open pull requests on (provider github), or a GitBook space it can open change requests in (provider gitbook). Also lists the account… - [Add a content source](https://docs.devtune.ai/api-reference/add-a-content-source.md): Maps a site URL prefix onto a GitHub repository folder (provider github, with installationId from getContentSources and repositoryId from listContentRepositories) or a GitBook space (provider gitbook, with connectionId, siteId and spaceId), with the checks Settings makes: GitHub must reach the repos… - [List content repositories](https://docs.devtune.ai/api-reference/list-content-repositories.md): Lists what addContentSource can map through one GitHub installation or one GitBook connection, read live as a click in Settings does: the repositories the installation reaches, or the spaces the connection's organization publishes and the address each is published at. Pass exactly one of installatio… - [Switch automatic pull requests](https://docs.devtune.ai/api-reference/switch-automatic-pull-requests.md): Turns automatic pull requests on or off for a content source, as the switch in Settings does: when on, each new draft of a page under the source opens a pull request or change request. Turning on needs content pull requests (403 content_pull_requests_unavailable); turning off does not, so it works a… - [Remove a content source](https://docs.devtune.ai/api-reference/remove-a-content-source.md): Removes a content source, as Remove does in Settings: drafts for pages under it stop opening pull requests or change requests, pull requests already open stay open, and its urlPrefix can be added again. Needs no content pull requests, and works on a paused project. A source already removed, or not i… - [Pause or resume a project](https://docs.devtune.ai/api-reference/pause-or-resume-a-project.md): Pauses or resumes the project. A paused project keeps its history readable but runs nothing new, frees its paid project slot, and refuses other setup changes until it is resumed, except disconnecting GA4 and changing content sources. Pausing needs the Agency plan (403 with reason agency_only). Resum… - [Get the Agent Picks catalogue](https://docs.devtune.ai/api-reference/get-the-agent-picks-catalogue.md): Lists what the setup wizard in the app offers for saveAgentPicksSetup and saveAgentPicksJob: published categories with their active tools, the test apps agents start in (one with category_ids only for those categories), and the agents with the models each accepts and whether the account’s Agent Pick… - [Get the AI platform catalogue](https://docs.devtune.ai/api-reference/get-the-ai-platform-catalogue.md): Lists the AI platforms DevTune probes, in display order, with whether the account’s plan includes each. Takes no query parameters. Requires the `setup.read` scope. - [List topics](https://docs.devtune.ai/api-reference/list-topics.md): Pages the project's topics with their names and descriptions. Pass a topic id as topicId to listPrompts, getVisibilitySummary or getCompetitivePosition. Only active topics are listed unless includeInactive is true; deleted topics are never listed. Ordered by id. Pass nextCursor back as cursor with i… - [Create a topic](https://docs.devtune.ai/api-reference/create-a-topic.md): Creates a topic in this project to group prompts. A topic with the same name returns 409, so a retry after a lost response returns 409 once the first call succeeded: list the topics to find it. Requires the `prompts.write` scope. - [Update a topic](https://docs.devtune.ai/api-reference/update-a-topic.md): Renames a topic, changes its description, or switches it on or off. A switched-off topic's prompts stop running and keep their own on/off state. A topic outside this project returns 404. Requires the `prompts.write` scope. - [Delete a topic](https://docs.devtune.ai/api-reference/delete-a-topic.md): Deletes a topic and every prompt in it. They stop running and free their allowance slots at once; history is kept, and restoring one of its prompts with restorePrompt brings the topic back. A topic that is already deleted, or outside this project, returns 404. Requires the `prompts.write` scope. - [List tracked prompts](https://docs.devtune.ai/api-reference/list-tracked-prompts.md): Pages the project's tracked prompts with their exact text, topic and lifecycle. Pass a prompt id as promptId to createPromptAction, listPromptActions or listVisibilityResponseHistory. By default only prompts that are on and whose topic is on are listed; includeInactive adds switched-off prompts and… - [Create a tracked prompt](https://docs.devtune.ai/api-reference/create-a-tracked-prompt.md): Tracks a new prompt in a topic of this project, switched on. It counts against the prompt allowance: over the allowance the request returns 409 with reason prompt_limit_reached and the allowance in the message. Text already tracked returns 409; a highly similar prompt returns 409 with reason similar… - [Update a tracked prompt](https://docs.devtune.ai/api-reference/update-a-tracked-prompt.md): Changes a prompt's text, moves it to another topic in this project, or switches it on or off. Switching it on counts against the prompt allowance and returns 409 with reason prompt_limit_reached when it is full; send isActive true on its own, without promptText or topicId. Pass expectedActive to ref… - [Delete a tracked prompt](https://docs.devtune.ai/api-reference/delete-a-tracked-prompt.md): Deletes a prompt. It stops running and frees its allowance slot at once; its history is kept and restorePrompt brings it back. Pass your own requestId to retry safely: the same requestId returns the recorded result. A prompt that is already deleted, or outside this project, returns 404. Requires the… - [Restore a deleted prompt](https://docs.devtune.ai/api-reference/restore-a-deleted-prompt.md): Brings back a deleted prompt, switched off; switch it on with updatePrompt. If its topic was deleted with it, the topic comes back too. Restoring a prompt that is not deleted changes nothing and returns alreadyRestored true. Returns 409 when the same text is tracked again or its topic was deleted on… - [Get prompt-set analysis](https://docs.devtune.ai/api-reference/read-prompt-set-analysis.md): Read the latest or selected analysis, paginated assessments and proposals, latest completed run identity, and credit state. Reads never reserve or settle credits or apply proposals. Pass pagination.nextCursor as cursor for the next page; the cursor stays on the run the first page read. You can also… - [Start or retry prompt-set analysis](https://docs.devtune.ai/api-reference/start-or-retry-prompt-set-analysis.md): Spends credits under the actions.write scope. Obtain human approval of the current estimate before calling. Requires an API key whose creator still has project management permission; this endpoint does not accept OAuth tokens. Over MCP, devtune_start_prompt_set_analysis also works with an OAuth conn… - [Review a prompt-set proposal](https://docs.devtune.ai/api-reference/review-a-prompt-set-proposal.md): Show the proposal and obtain explicit human approval of apply, undo, dismiss or restore. Requires an API key whose creator still has project management permission. Uses the same quota, stale-state and concurrency checks as in-app review. No analysis credit charge. Restoration returns a dismissed pro… - [List AI response history](https://docs.devtune.ai/api-reference/list-ai-execution-history.md): Lists the AI answer runs DevTune still holds, with every status by default, newest first, in a fixed 30- or 90-day window ending when the first page is read. Each item is metadata only, without answer text or counts; read one run with getVisibilityResponse for its answer evidence. Use this list to f… - [List recent AI responses](https://docs.devtune.ai/api-reference/list-recent-ai-responses.md): Lists the latest 20 successful executions, optionally filtered by prompt or platform. Use an executionId with the response detail endpoint to inspect retained evidence. This operational list is not a measurement sample; use visibility summary for rates. Requires the `visibility.read` scope. - [Get a retained AI response](https://docs.devtune.ai/api-reference/get-a-retained-ai-response.md): Reads one execution as evidence, including captured sponsored ads. Get executionId from the response history or recent responses listing. This is an entity detail read, not a measurement query. Use visibility summary for advertising rates. retained=false means no nonempty public answer or ad evidenc… - [Get visibility summary metrics](https://docs.devtune.ai/api-reference/get-visibility-summary-metrics.md): Returns share of voice, presence rate, brand mentions, and citation statistics. The default scope includes only platforms currently active for the project. Requires the `visibility.read` scope. - [Get brand sentiment](https://docs.devtune.ai/api-reference/get-brand-sentiment.md): Returns the sentiment of AI answers towards the project's primary brand over a fixed window of completed UTC days ending yesterday: overall and for each day, across every platform. Each answer that mentions the primary brand is scored from -1 to 1 for its stance towards that brand. A language model… - [Get probe health per AI platform](https://docs.devtune.ai/api-reference/get-probe-health-per-ai-platform.md): Returns, for each AI platform scheduled probes run on for the project, the latest scheduled run's prompt outcomes, when the platform last succeeded and any open platform outage DevTune has published. Check it before retrying a Visibility read that looks stale. Runs and successes are read from the la… - [Get visibility diff](https://docs.devtune.ai/api-reference/get-visibility-diff.md): Compares the current visibility window with the previous same-length window. comparisons says whether each share moved clearly; report a change only when comparison is up or down. Requires the `visibility.read` scope. - [Get project timeline events](https://docs.devtune.ai/api-reference/get-project-timeline-events.md): Returns project changes and sourced global facts as context for interpreting Visibility trends. Global facts are limited to all-platform events or platforms enabled for the project. Events do not imply a causal link to metric movements. Requires the `visibility.read` scope. - [Get competitive positioning data](https://docs.devtune.ai/api-reference/get-competitive-positioning-data.md): Returns citation comparison between primary, competitor, and other sources over time for a fixed rolling window. The default includes only currently active project platforms. Requires the `competitive.read` scope. - [Get aggregate citation statistics](https://docs.devtune.ai/api-reference/get-aggregate-citation-statistics.md): Returns aggregate citation totals for a fixed 30- or 90-day window. Coverage tells data that is not ready yet apart from a measured zero. Use top-pages and top-domains for citation lists. Requires the `citations.read` scope. - [Get top citation domains](https://docs.devtune.ai/api-reference/get-top-citation-domains.md): Returns top citation domains for a fixed 30- or 90-day window. Coverage tells data that is not ready yet apart from a measured zero. Pass topicId, promptId or platform to count only the citations from that topic, tracked prompt or AI platform; filters combine and apply before ranking and paging. A f… - [Get top citation pages](https://docs.devtune.ai/api-reference/get-top-citation-pages.md): Returns top citation pages for a fixed 30- or 90-day window. Coverage tells data that is not ready yet apart from a measured zero. Pass topicId, promptId or platform to count only the citations from that topic, tracked prompt or AI platform; filters combine and apply before ranking and paging. A fil… - [List Lookalike sites](https://docs.devtune.ai/api-reference/list-lookalike-sites.md): Lists the hosts AI answers cited in a 7, 30 or 90 day window whose domain imitates one of the project’s own domains, most cited first, with how many answers cited each out of the answers checked, overall and per platform. Requires the `citations.read` scope. - [List Lookalike site citations](https://docs.devtune.ai/api-reference/list-lookalike-site-citations.md): Lists the answers that cited one lookalike host, newest run day first, with the cited URL, how the answer framed the link and the text around it. Returns 404 for a host outside the project. Requires the `citations.read` scope. - [Get positioning match](https://docs.devtune.ai/api-reference/get-positioning-match.md): Returns positioning counts, resolved brand names, daily counts and share_of_statement. Share is the primary brand count divided by brand_counts plus another_product, excluding several and nobody. Below ten named-product assignments the share is null. Each statement also carries prior_status and, whe… - [List positioning statements](https://docs.devtune.ai/api-reference/list-positioning-statements.md): Returns id, kind, statement, relevant_intents, status and created_at for active and archived positioning statements. Requires the `positioning.read` scope. - [List positioning match answers](https://docs.devtune.ai/api-reference/list-positioning-match-answers.md): Lists the AI answers behind one positioning match cell, newest first: accepted verdicts for the statement and assignee over the same window, filters and current competitor set as getPositioning, so the number of answers equals that cell. A statement that is a fact, proposed, dismissed or not in this… - [Get Accuracy](https://docs.devtune.ai/api-reference/get-accuracy.md): Returns how the claims AI answers made about the product compare with its own site over a 30 or 90 day window: claim groups by outcome and the match rate, the previous window when it compares, the weekly trend, the per-platform split and check coverage. Each claim group has one outcome, the most sev… - [List Accuracy claims](https://docs.devtune.ai/api-reference/list-accuracy-claims.md): Lists the claim groups with one outcome over the same window and filters as getAccuracy, most answers first, then most recently seen. Each item carries the latest answer with that outcome, the site page and any other page, and the team’s active decision. Requires the `accuracy.read` scope. - [Get Accuracy claim](https://docs.devtune.ai/api-reference/get-accuracy-claim.md): Returns one claim group over the same window and filters as getAccuracy: its outcome, platforms and answers, the latest answer with its pages and the evidence it was judged against, up to 20 other answers, the brand’s own pages the answers cited, and the team’s active decision. A merged group opens… - [List Accuracy contradictions](https://docs.devtune.ai/api-reference/list-accuracy-contradictions.md): Lists contradictions once per site sentence they conflict with, over the same window and filters as getAccuracy, most answers first, then most recently seen: the rows of the Accuracy tab’s contradictions table. Each row carries the sentence and its page, how many AI wordings (claim groups) contradic… - [Get Accuracy contradiction](https://docs.devtune.ai/api-reference/get-accuracy-contradiction.md): Returns one contradiction row over the same window and filters as getAccuracy: the site sentence and its page, the row’s counts, platforms and open action, and every AI wording that contradicts it, active first, with its latest contradicting answer and the decision that dismissed it. A row with no w… - [List Accuracy decisions](https://docs.devtune.ai/api-reference/list-accuracy-decisions.md): Lists the decisions the team made on Accuracy claims, active and undone, newest first. Requires the `accuracy.read` scope. - [List content gaps](https://docs.devtune.ai/api-reference/list-content-gaps.md): Lists topic, source and brand content gaps over a fixed 30- or 90-day window: targets where competitors are cited, with citations split into primary and competitor, average position and your share. Your own sources and brand are excluded, as are topics you hold at least half the citations on, becaus… - [Get content gap evidence](https://docs.devtune.ai/api-reference/get-content-gap-evidence.md): Returns the tracked prompts and cited pages behind one content gap from List content gaps. Pass its gapType and targetKey with the same windowDays. Evidence is published with the gap, so it covers the same fixed window and asOfDate. A source gap matches its URL, a brand gap the pages cited on its do… - [List recent answer fetches](https://docs.devtune.ai/api-reference/list-recent-answer-fetches.md): Returns up to 20 answer-fetch hourly records, newest hour first with descending id as the tie-break. The selected UTC calendar window includes today up to the server request time. Each item is one hourly record, not an individual request or a window total. An empty list means no records were found;… - [Get visits per answer fetch](https://docs.devtune.ai/api-reference/get-visits-per-answer-fetch.md): Returns AI-referred visits per answer fetch for each page and for the whole project, over completed UTC days ending the day before yesterday by default. An answer fetch is an AI system retrieving the page while building an answer; it is not a citation. A page is covered when its visits come from GA4… - [List AI referrals](https://docs.devtune.ai/api-reference/list-ai-referrals.md): Returns paginated AI-referral landing pages, overview and the complete selected-provider engine breakdown for the same fixed reporting dates. Sessions are not people; conversion events use the selected provider and conversion definitions. Nullable provider measurements remain unavailable. Known coun… - [Get coding-agent answer fetches](https://docs.devtune.ai/api-reference/get-coding-agent-answer-fetches.md): Returns answer fetches by coding agents over a fixed rolling window, per agent and platform since countedFrom: the coding-agent counts of the summary's codingAgents, without its totals of all answer fetches. Every Agent Picks plan can read it, with 404 Demand; the other traffic reads need an AI Sear… - [Get AI Traffic sensors](https://docs.devtune.ai/api-reference/get-ai-traffic-sensors.md): Returns the status of six AI Traffic sensors (edge middleware, snippet, Cloudflare, Vercel drain, Search Console and GA4): whether each is connected, the measurement verdict for a set-up sensor, when DevTune last accepted a measurement from it, and why its connection is broken. Lists Cloudflare, Sea… - [Disconnect GA4](https://docs.devtune.ai/api-reference/disconnect-ga4.md): Disconnects one Google Analytics 4 connection, as Disconnect does in AI Traffic settings: DevTune stops syncing it, deletes its stored credential and attempts to revoke the Google grant; a failed revoke does not block the disconnect. Repeating it on a disconnected connection succeeds. Reconnecting n… - [Rotate the edge ingest key](https://docs.devtune.ai/api-reference/rotate-the-edge-ingest-key.md): Issues a new ingest key for the edge middleware and returns it. The previous key keeps working for 24 hours after the rotate, or until revokePreviousTrafficIngestKey. The key is shared by every site that sends traffic to this project, so a person deploys the new key to all of them before anyone revo… - [Revoke the previous edge ingest key](https://docs.devtune.ai/api-reference/revoke-the-previous-edge-ingest-key.md): Stops the previous ingest key working now, so the edge middleware accepts only the current key: revoke it only once every site that sends to this project uses the current key. revoked is false when no previous key was still working, so repeating it is safe. Needs AI Traffic Events in the plan, as AI… - [Save a Cloudflare source](https://docs.devtune.ai/api-reference/save-a-cloudflare-source.md): Saves a Cloudflare API token for a zone, as Save does in AI Traffic settings: creates the source, or replaces the token of the source for that zone and sets it active. When the source needed a new token, its hostnames waiting on it (status needs_reauth) become active again; paused hostnames and host… - [Save a Cloudflare hostname](https://docs.devtune.ai/api-reference/save-a-cloudflare-hostname.md): Starts the Cloudflare log pull for one of the project’s own primary tracked domains, or changes its source, hostname or path prefix, as Save does in AI Traffic settings. The hostname must be the tracked domain’s host and the path prefix must sit within its path. A saved hostname is active. The sourc… - [Pause or resume a Cloudflare hostname](https://docs.devtune.ai/api-reference/pause-or-resume-a-cloudflare-hostname.md): Pauses or resumes the log pull for one Cloudflare hostname, as Pause and Resume do in AI Traffic settings. Resuming a broken hostname puts it back in the next pull. Both need AI Traffic Events in the plan, as the settings page does. Resuming returns 409 while the source needs a new token; replace it… - [Delete a Cloudflare hostname](https://docs.devtune.ai/api-reference/delete-a-cloudflare-hostname.md): Stops the log pull for one Cloudflare hostname and deletes it, with the AI Traffic counts DevTune pulled for it, as Delete does in AI Traffic settings. Deleting the last hostname on a source also deletes the source and its stored token. Needs AI Traffic Events in the plan, as the settings page does.… - [Disconnect Search Console](https://docs.devtune.ai/api-reference/disconnect-search-console.md): Disconnects one Google Search Console connection, as Disconnect does in AI Traffic settings: DevTune stops syncing it and deletes its stored credential, and keeps the data already synced. Repeating it on a disconnected connection succeeds. Reconnecting needs a person to open the search_console conne… - [List AI referral hosts](https://docs.devtune.ai/api-reference/list-ai-referral-hosts.md): Lists the project's AI referral hosts in hostname order: its owned domains and every host a server-side sensor (edge middleware, Cloudflare or a Vercel log drain) has reported, plus any excluded host outside that set. A host only the tracking snippet reports is not listed. Project-wide referrals cou… - [Get 404 Demand](https://docs.devtune.ai/api-reference/get-404-demand.md): Returns project-scoped content-intent and suspected-noise 404 clusters for one published window, including explicit degraded state, machine-traffic sensor coverage and the days the window was measured. The 30-day window, the default, is the one opportunities and actions use. The 7-day window is mine… - [Get AI-correlated Direct lift](https://docs.devtune.ai/api-reference/get-ai-correlated-direct-lift.md): Returns the project's Direct traffic totals from GA4, each measured page's actual and baseline Direct traffic per day, and how each page-window is classified, for a fixed 7, 30, or 90 day request. AI-correlated lift is correlated evidence, never attributed traffic, and is never summed with measured… - [Get traffic summary](https://docs.devtune.ai/api-reference/get-traffic-summary.md): Daily counts are null on UTC days without machine-sensor coverage and 0 on covered days with no events. Window totals sum covered days only and are null when no day was covered, or when the window is only partly covered and the sum is 0. Each total has a Status sibling: measured when its cove… - [Get Search Console and analytics totals](https://docs.devtune.ai/api-reference/get-search-rollup-totals.md): Returns fixed-window totals for Google Search Console (gsc), legacy GA4 (ga4) or the selected analytics provider (analytics). Read the returned source window dates instead of assuming the requested window: Search Console days are Pacific-time days ending three days before today, and coverage is limi… - [Get a page's Search Console queries](https://docs.devtune.ai/api-reference/get-a-pages-search-console-queries.md): Returns the Search Console queries one page ranks for in a fixed 30- or 90-day window, filtered to a position band and an impression floor, ordered by impressions. The defaults (positions 8 to 20, at least 20 impressions, 90 days) select queries where the page already ranks and an edit to the page c… - [Get traffic by platform](https://docs.devtune.ai/api-reference/get-traffic-by-platform.md): Returns LLM traffic grouped by AI platform, traffic type and bot class, with event counts, bot-identity verdict and unknown-reason sums, and percentages. Requires the `traffic.read` scope. - [Get adoption metrics](https://docs.devtune.ai/api-reference/get-adoption-metrics.md): Returns npm downloads and GitHub stars trends for project domains. Requires the `adoption.read` scope. - [Get Audit summary](https://docs.devtune.ai/api-reference/get-audit-summary.md): Returns completed Audit summaries and independently bounded trend calendars. date belongs to UTC citations and answer fetches; clicksDate and priorClicksDate belong to Pacific GSC source windows ending at D-3. Read GSC coverage and bounds from searchConsole. Unverified days are null. Requires the `a… - [Get the extracted section for an Audit finding](https://docs.devtune.ai/api-reference/get-stored-extraction-for-a-section-finding.md): Returns the page excerpt a section finding was based on: the section and up to 6,000 characters of nearby Markdown, with extractedAt for when the page was read. Pass domain, pagePath, contentHash, inputHash and locator exactly as the finding returned them. status is available when the excerpt is ret… - [Acknowledge or reconsider a retrieval restriction](https://docs.devtune.ai/api-reference/acknowledge-or-reconsider-a-retrieval-restriction.md): Records customer intent for exact URLs displayed in the stored audit finding. confirmedIntent: true attests that the customer confirmed it; DevTune checks only that it is true. Preserves restrictions, evidence and score. Requires project management permission. Requires the `actions.write` scope. - [List Audit findings](https://docs.devtune.ai/api-reference/list-audit-findings.md): Returns the findings that Audit shows. Arbitrary date ranges are not accepted. affectedUrls is a capped sample; affectedPageCount reports the uncapped population the audit measured, in the unit named by affectedScope, and is null when a finding covers no countable population. Findings and sectionAss… - [Dismiss or restore an Audit finding](https://docs.devtune.ai/api-reference/dismiss-or-restore-an-audit-finding.md): Dismisses one finding of a completed audit run in this project, or restores a dismissed one to open, as Dismiss and Restore on the Audit page do. Take auditRunId and findingId from List Audit findings; pass includeDismissed=true there to find dismissed ones. The change is made as the member who crea… - [Create an action from an Audit finding](https://docs.devtune.ai/api-reference/create-an-action-from-an-audit-finding.md): Creates an accepted action from one finding of a completed audit run in this project, as Create action on the Audit page does: the same action, finding evidence and target pages. Take auditRunId and findingId from List Audit findings. title defaults to the finding's remediation summary. The action i… - [Get Google eligibility evidence](https://docs.devtune.ai/api-reference/get-google-eligibility-evidence.md): Returns whether Google can index your pages, from Search Console URL Inspection: each eligibility check, status changes confirmed by two inspections in a row, and sitemap submission state, for a fixed 7, 30, or 90 day window. Evidence is dated from successful measurements. Detail lists are capped wh… - [List page metrics](https://docs.devtune.ai/api-reference/list-unified-page-metrics.md): Returns per-page metrics for a fixed 7-, 30- or 90-day window. Page rows and summary cover content pages only; searchConsole holds the whole-project Search Console totals across all pages, with current and prior coverage and source dates. The last included Search Console day is three days before tod… - [List page guard-rail findings](https://docs.devtune.ai/api-reference/list-page-guard-rail-findings.md): Returns the page guard-rail findings that Audit shows. Arbitrary date ranges are not accepted. Pass url to return only the rows page protection checks for that page; any of them with isUserLocked true means the page is locked. An empty filtered list means DevTune holds no guard-rail row for the page… - [Lock a page](https://docs.devtune.ai/api-reference/lock-a-page.md): Locks a page on one of the project's own tracked domains, so actions, drafts and agents steer around it. The page does not need to be crawled yet. A domain that is not one of the project's own tracked domains returns 404. Locking a page that is already locked keeps it locked, so a retry is safe. Req… - [Unlock a page](https://docs.devtune.ai/api-reference/unlock-a-page.md): Unlocks a locked page, so suggestions may change it again. A page that is not locked returns 404. Requires the `setup.write` scope. - [List actions for a prompt](https://docs.devtune.ai/api-reference/list-actions-for-a-prompt.md): Returns up to 100 existing and suggested actions explicitly linked by stored prompt evidence, with status and brief readiness. hasMore indicates truncated results. Requires the `actions.read` scope. - [Create an action from a prompt](https://docs.devtune.ai/api-reference/create-an-action-from-a-prompt.md): Creates an action for the prompt: with targetUrl, to refresh that page; without it, to create new content under title. An existing action for the same page, or for new content with the same title, is returned instead of a second one. Saves the prompt's evidence from the last 30 days. generateBrief d… - [List actions](https://docs.devtune.ai/api-reference/list-actions.md): Returns the project's actions. Each is suggested (DevTune proposed it and nobody has decided), accepted (your team wants it done), live (the page is published; its measurement is running or complete) or archived (taken off the queue). Filter with status; without it, every state is returned, suggesti… - [Update an action](https://docs.devtune.ai/api-reference/update-an-action.md): Changes the title, owner, channels or description of an accepted or live action. Restore an archived action first. DevTune's suggestions, including dismissed and snoozed ones, keep their details and have no owner; accept one before changing it. assignedTo is the user id or sign-in email of a current… - [Change an action status](https://docs.devtune.ai/api-reference/change-an-action-status.md): Moves an action through its lifecycle, as the member who created the API key; they must be a project member with permission to manage actions (403 otherwise, including for a key with no creator). Transitions come in pairs. accept moves a suggestion to accepted and unaccept returns it to the suggesti… - [Get an action spot check](https://docs.devtune.ai/api-reference/get-an-action-spot-check.md): Returns an action's spot check, or spotCheck null before one starts: its status, when it was requested and completed, and one result per prompt and platform with whether the answer mentioned the brand, whether it cited the brand's site, and up to 20 of its citations. A platform that could not answer… - [Start an action spot check](https://docs.devtune.ai/api-reference/start-an-action-spot-check.md): Starts a live action's spot check, as the member who created the API key or authorized the OAuth connection; they must be a project member with permission to manage actions (403 otherwise). DevTune asks the action's own prompts once, its trigger prompts first and up to 12 of them, on each AI platfor… - [Split a merged suggestion](https://docs.devtune.ai/api-reference/split-a-merged-suggestion.md): Splits a suggestion DevTune's duplicate check merged into a decided action back out, as the member who created the API key or authorized the OAuth connection; they must be a project member with permission to manage actions (403 otherwise). Pass a normalizedLabel from an action's context.mergedSugges… - [List action edits](https://docs.devtune.ai/api-reference/list-action-edits.md): Lists who changed an action's title, owner, channels, description or published pages, newest first, one event per changed field and one per published page added or removed. Owners are user ids, channels are arrays and pages are URLs. A member who made an edit is named while they are still on the acc… - [Get action content draft](https://docs.devtune.ai/api-reference/get-action-content-draft.md): Returns an action's content draft for accounts with content drafts: the draft the action takes now, the models a run can use with their estimates, the run in flight, and the current draft with up to 10 earlier versions, each with its markdown, change list and fact check. Refused with 403 when conten… - [Start an action content draft](https://docs.devtune.ai/api-reference/start-an-action-content-draft.md): Starts a content-draft run for the action with the same checks as Generate draft in the app: the Page Drafter for a create action, the Section Editor for a refresh action. A refusal carries its reason: 409 for not_content_action, read_only, brief_not_ready, brief_outdated (the brief predates the cur… - [Create action draft pull request](https://docs.devtune.ai/api-reference/create-action-draft-pull-request.md): Opens a pull request on the customer's content repo through the DevTune GitHub App, or a change request in their GitBook space, from the action's current content draft, or writes a newer draft to the open one. The page belongs to the content source with the longest matching URL prefix, and the sourc… - [Close action draft pull request](https://docs.devtune.ai/api-reference/close-action-draft-pull-request.md): Closes the action's open pull request in GitHub and deletes DevTune's branch, or archives its GitBook change request. Works even when content pull requests are no longer enabled. The draft stays. A refusal is a 409 with its reason: no_pull_request, pull_request_merged or source_needs_reconnect (reco… - [Get action brief](https://docs.devtune.ai/api-reference/get-action-brief.md): Returns the stored brief status for an action and, when a brief is ready, the brief. A brief written to the brief contract (briefFormat 2) is returned as Markdown, brief first and its Evidence section after, and as the structured brief and evidence objects. A briefFormat 1 brief returns Markdown onl… - [Generate action brief](https://docs.devtune.ai/api-reference/generate-action-brief.md): Generates an execution-ready brief for an action DevTune suggested, including one created from a prompt, or reuses the one it has. It spends credits under the actions.write scope: a new brief reserves 2 credits and settles at 1–2 credits, depending on how much research it needs; reusing a ready or r… - [List detected content changes](https://docs.devtune.ai/api-reference/list-detected-content-changes.md): Lists unclaimed and claimed changes with measured-action IDs and current manual claim tokens. Deltas compare fixed 14-day windows. Path filtering applies to list totals; summary covers the entire selected window. Missing evidence remains null. Requires the `outcomes.read` scope. - [Get published-page tracking status](https://docs.devtune.ai/api-reference/read-published-page-tracking-status.md): Read-only operational status for saved URLs. Omit actionId for project-wide capacity and allowance status. At a plan limit the remedy is to upgrade; the subscription plan determines the allowance. Report these states as unmeasured. Requires the `outcomes.read` scope. - [Attach a published URL to a completed action](https://docs.devtune.ai/api-reference/attach-a-published-url-to-a-completed-action.md): Measurement starts at the action completion date, which is the declared went-live date. Crawling and page-tracking allowance do not gate measurement. Send every published URL the action should keep in publishedUrls: any you leave out is removed. Existing anchored measurements are preserved; changeEv… - [Claim a content change](https://docs.devtune.ai/api-reference/claim-a-content-change.md): Creates a completed measured action, or restores the same archived action and measurement on re-claim. Repeated calls return the original IDs and do not overwrite action copy. The claim returns at once; the change list shows measurementUpdating while the measurement catches up. Returns 409 if the pa… - [Unclaim a content change](https://docs.devtune.ai/api-reference/unclaim-a-content-change.md): Reverses only an unchanged manual claim using its current claimToken. Archives the action and excludes the retained measurement from totals. Preserves history. Returns 409 for stale tokens, edited or linked work, automatic attribution, or claims without reversible history. Requires the `actions.writ… - [List measured results](https://docs.devtune.ai/api-reference/list-measured-results.md): Returns the Results ledger and scoreboard, plus analytics-attributed conversions for fixed 30 and 90 day windows. An unmeasured conversion window returns null. Only the fixed 7, 30, 90, and 365 day ledger windows are accepted. Requires the `outcomes.read` scope. - [Get the intelligence forecast](https://docs.devtune.ai/api-reference/get-the-intelligence-forecast.md): Projects primary citations and presence rate forward from a straight-line fit over the daily figures in the window. It is a trend extrapolation, not a measured result. Presence-rate values are ratios from 0 to 1 and are null when no response in the window has usable citation data. With no data in th… - [Get what-works recommendations](https://docs.devtune.ai/api-reference/get-what-works-recommendations.md): Returns which channels have shipped improvements, from Results whose 14-day before/after comparison has closed as improved on citations or answer fetches. The window selects when those results became available. Open, mixed, no_clear_change, declined and unmeasured results are excluded. sampleSize co… - [List managed agents](https://docs.devtune.ai/api-reference/list-managed-agents.md): Returns roster and active custom agents with display fields and, for each, the evidence verdict a run would start from. With archived=true, returns only archived custom agents instead, each with archivedAt, so you can find the agentKey to pass to restoreCustomAgent. Requires the `agents.read` scope. - [Create a custom agent](https://docs.devtune.ai/api-reference/create-a-custom-agent.md): Creates a custom agent, as Create agent does on the Agents page, with the same validation; a model that is not available now is refused with 400. The agent starts with no schedule. The response carries its agentKey. A retry creates a second agent. The OAuth member, or the member who created the API… - [List managed-agent runs](https://docs.devtune.ai/api-reference/list-managed-agent-runs.md): Returns project-scoped managed-agent run status and definition display fields. Draft output is available only from the single-run endpoint after completion. Requires the `agents.read` scope. - [Get a managed-agent run](https://docs.devtune.ai/api-reference/get-a-managed-agent-run.md): Returns run status and the complete action, brief or report output only after the run reaches a completed review state. Content-draft output remains null. Requires the `agents.read` scope. - [Review a managed-agent draft](https://docs.devtune.ai/api-reference/review-a-managed-agent-draft.md): Approves or dismisses the draft of a run awaiting review (status draft_ready), as Approve and Dismiss do in the app. Approving an action or brief draft adds it to Actions, and approving a report keeps it. Dismissing discards the draft and any hidden action it created. A Company Context Refresher pro… - [Send a managed-agent run a follow-up](https://docs.devtune.ai/api-reference/send-a-managed-agent-run-a-follow-up.md): Queues a follow-up of a completed run (status draft_ready, approved or dismissed), as Run follow-up does in the app: a new metered run of the same agent version, grounded in the prior draft and the instructions. Credits are reserved and settled as for triggerManagedAgentRun, and with too few credits… - [Update a custom agent](https://docs.devtune.ai/api-reference/update-a-custom-agent.md): Changes a custom agent's name, description, instructions, model, quick actions or tools. Send only the fields to change; the result must still be a valid agent by the Agents page's rules. If the agent's model is no longer available, the request is refused with 400 and reason model_unavailable until… - [Archive a custom agent](https://docs.devtune.ai/api-reference/archive-a-custom-agent.md): Archives a custom agent, as Archive does on the Agents page. The agent leaves listManagedAgents, its schedule turns off, and runs, edits and schedules for it are refused with 409 and reason agent_archived until it is restored. A run already started finishes. Its past runs, drafts and briefs stay rea… - [Restore an archived custom agent](https://docs.devtune.ai/api-reference/restore-an-archived-custom-agent.md): Restores an archived custom agent, as Restore does on the Agents page. It returns to listManagedAgents with its schedule off; turn the schedule back on with setAgentSchedule. Restoring an agent that is not archived changes nothing. Built-in agents are never archived (400, reason built_in_agent). A c… - [Set an agent schedule](https://docs.devtune.ai/api-reference/set-an-agent-schedule.md): Turns an agent's schedule on or changes it (autonomyLevel scheduled, with schedule and modelId), or turns it off (autonomyLevel manual, without a schedule), as Schedule does on the Agents page, for built-in and custom agents. Other refusals, including invalid schedule inputs and unavailable models,… - [Turn off an agent schedule](https://docs.devtune.ai/api-reference/turn-off-an-agent-schedule.md): Turns an agent's schedule off. Future scheduled runs stop; a run already started continues. An agent with no current revision is already off. An archived agent can be turned off. Other inactive agents are refused with 409, reason agent_inactive and message: This agent is inactive, so its schedule ca… - [Trigger a managed-agent run](https://docs.devtune.ai/api-reference/trigger-a-managed-agent-run.md): Queues a manual run of the agent, which spends managed-agent credits. Repeating the request with the same requestIdempotencyKey returns that run as it stands, including a blocked or failed outcome, unless it is still queued without a credit reservation; only then is the retry checked like a new requ… - [List Dex conversation threads](https://docs.devtune.ai/api-reference/list-dex-conversation-threads.md): Lists your private Dex conversations in this project from recent history. Only the user attributed to the credential sees their conversations. truncated is true when older history is not included. creditTotalIsFinal is false while a charge is still settling or when older history for that thread is n… - [Start a Dex conversation](https://docs.devtune.ai/api-reference/start-a-dex-conversation.md): Starts a Dex conversation with your first message. Repeating the request with the same messageId returns the conversation it started instead of starting another. Dex runs spend managed-agent credits; with too few credits the request returns 402. The conversation is private to the member who created… - [Rename a Dex conversation thread](https://docs.devtune.ai/api-reference/rename-a-dex-conversation-thread.md): Renames one of your Dex conversations. Conversations are private to the member who created the API key, so another member's conversation returns 404. Renaming spends no credits, and sending the same title again changes nothing. Returns 402 when the account doesn't include Dex. Requires the `agents.r… - [Delete your private Dex chat](https://docs.devtune.ai/api-reference/delete-your-private-dex-chat.md): Stops live work and removes the chat from its owner’s history. Billing and audit records are retained. The API key creator must own the chat. Requires the `agents.run` scope. - [List Dex conversation events](https://docs.devtune.ai/api-reference/read-dex-conversation-events.md): Returns one page of a Dex conversation: its replies and status events, oldest first, from recent history. Pass nextAfter from the previous page as after to read the next page; threadTruncated is true when older history is not included. Requires the `agents.read` scope. - [Send a Dex conversation message](https://docs.devtune.ai/api-reference/send-a-dex-conversation-message.md): Sends a message in one of your Dex conversations: it steers the task Dex is working on, or resumes a conversation that has gone idle. Resuming starts a Dex run, which spends managed-agent credits; with too few credits the request returns 402. This endpoint has the stricter rate limit that agent runs… - [List pending Dex specialist handoffs](https://docs.devtune.ai/api-reference/list-pending-dex-specialist-handoffs.md): Lists the specialist agent runs Dex has proposed in one of your conversations that are waiting for your decision and have not expired. estimateCredits is the estimated managed-agent credits the run would spend, or null when there is no estimate; expiresAt is when the proposal lapses. Confirm or canc… - [Confirm or cancel a Dex specialist handoff](https://docs.devtune.ai/api-reference/confirm-or-cancel-a-dex-specialist-handoff.md): Confirms or cancels a specialist agent run Dex proposed. confirm starts the run as the member who created the API key, and the run spends managed-agent credits; the handoff's estimateCredits is the estimate. The result status is queued, with the run's taskId, when the run started; not_started when i… - [Save or dismiss a Dex memory offer](https://docs.devtune.ai/api-reference/save-or-dismiss-a-dex-memory-offer.md): Saves or dismisses a memory Dex offered in one of your conversations. To save, send offeredText exactly as Dex offered it, with its offerId and taskId. Different text, or an unknown offerId, returns 404. A saved memory becomes a Library item that agents can use. Requires the `agents.run` scope. - [Update Agent Picks model settings](https://docs.devtune.ai/api-reference/update-agent-picks-model-settings.md): Save ordered agent/model pairs for paid tests on an existing setup. Duplicate pairs, unpriced or mismatched models and empty selections are refused. Every Agent Picks plan includes every agent. The starter test keeps its fixed models. The member who created the API key needs project management permi… - [Get Agent Picks summary](https://docs.devtune.ai/api-reference/get-agent-picks-summary.md): Returns which tool coding agents picked when asked for one, with each pick rate’s install rate beside it, from the project’s latest completed non-directional own test, otherwise its latest completed directional own test, otherwise its category’s public test: the copy cloned into the project, else th… - [List Agent Picks runs](https://docs.devtune.ai/api-reference/list-agent-picks-runs.md): List the setup’s run history, newest first by created_at then id, or the runs of the test test_id names, whichever setup it belongs to. Includes starter and paid runs. Public tests are not included. This lists runs, not rates; use the summary for rates. Requires Agent Picks. limit defaults to 50, ma… - [Get Agent Picks run evidence](https://docs.devtune.ai/api-reference/get-agent-picks-run-evidence.md): Inspect a run owned by the account, project and test together. Includes queued, running, failed, cancelled and completed runs. Request and job text are withheld; request_withheld and job_withheld distinguish withheld snapshots from absent ones; pick is null without a classification. Events are recor… - [Get Agent Picks run trace](https://docs.devtune.ai/api-reference/get-agent-picks-run-trace.md): Read one page of a project-owned run's console trace: redacted prompt, setup, narration, tool calls and results, file edits and errors, ordered by attempt then seq. An attempt's lines are available once that attempt finishes, so a running attempt has none yet; an attempt whose transcript could not b… - [Quote an Agent Picks test](https://docs.devtune.ai/api-reference/quote-an-agent-picks-test.md): Quote a paid test, an Action re-test or one run of a schedule in credits, without spending anything. Pass test options (preset quick or cell, runs_per_cell, scope) for a test; action_id alone for a re-test of a completed or live managed Action (only those; at most 40 runs; a self-run Action is refus… - [Start an Agent Picks test](https://docs.devtune.ai/api-reference/start-an-agent-picks-test.md): Start a quoted paid test or re-test. Pass quote_token from the quote and max_credits, the most this test may reserve, in whole credits. The start re-quotes and reserves the fresh high estimate; settlement never charges more, and unused credits return when the test ends. A retry with the same quote t… - [Cancel an Agent Picks test](https://docs.devtune.ai/api-reference/cancel-an-agent-picks-test.md): Cancel a queued or running paid test or re-test. Queued runs are cancelled at once; running runs keep their reservation until they settle, so credits may return later: refunded_micros and released_at on the test read show when. A repeat returns status cancelled again. Starter tests cannot be cancell… - [Create an Agent Picks self-run upload](https://docs.devtune.ai/api-reference/create-an-agent-picks-self-run-upload.md): Open an upload for a bundle the agentpicks CLI wrote on the customer's own machine. The body is the bundle's manifest.json, which declares each artifact's name, bytes and sha256. The response gives one signed Storage address per declared artifact: PUT the artifact's bytes to its url within two hours… - [Get an Agent Picks self-run upload](https://docs.devtune.ai/api-reference/get-an-agent-picks-self-run-upload.md): Read one self-run upload: open, completed or expired, and once completed the recorded run with its status and its analysis: pending, analysed, or failed with why. An open upload past its 24 hours reads as expired, as does one that held an artifact of the wrong size. A deleted upload, or another proj… - [Delete an Agent Picks self-run upload](https://docs.devtune.ai/api-reference/delete-an-agent-picks-self-run-upload.md): Delete an upload and its artifacts. Its artifacts are removed by a cleanup sweep that runs every 15 minutes, no sooner than 2 hours 15 minutes after the upload was last created or retried, when its addresses have lapsed; until then the run id can't be created again. For a completed upload this also… - [Delete an Agent Picks self-run series](https://docs.devtune.ai/api-reference/delete-an-agent-picks-self-run-series.md): Delete a self-run series in one call: every upload of it not already deleted (completed, open or expired) and every month of its recorded runs, so its self-run tests go and the Actions they fed are recomputed. Returns how many uploads and runs it deleted. Artifacts are removed by a cleanup sweep tha… - [Complete an Agent Picks self-run upload](https://docs.devtune.ai/api-reference/complete-an-agent-picks-self-run-upload.md): Complete an open upload once every artifact its manifest declares is stored at its declared bytes. The run is recorded as a self-run, completed and awaiting analysis, in its series' self-run test for the month it started, and its analysis is requested; run.analysis reads analysed once it has an outc… - [Get Agent Picks schedule](https://docs.devtune.ai/api-reference/get-agent-picks-schedule.md): Read cadence, next UTC run, the newest skipped period with its reason and recovery action, and the current paid-test estimate per run and for the next calendar month against included monthly-grant credits from active subscriptions only, zero while trialing. Quote failures return estimate null. last_… - [Enable Agent Picks schedule](https://docs.devtune.ai/api-reference/enable-agent-picks-schedule.md): Enable or change the cadence: daily at midnight UTC, weekly on Monday, or monthly on the first. Quote with schedule set to the same cadence first, then pass cadence, its quote_token and max_credits, the most one scheduled run may reserve. A quote_token works once: reusing it is refused as 409 quote_… - [Turn off Agent Picks schedule](https://docs.devtune.ai/api-reference/turn-off-agent-picks-schedule.md): Turn the schedule off. Future scheduled tests stop; a test already queued or running continues until cancelled. Works while the account is paused. The member who created the API key needs project management permission and edit rights on the project; otherwise 403. This endpoint takes an API key, not… - [Get Agent Picks test cells](https://docs.devtune.ai/api-reference/get-agent-picks-test-cells.md): Read per-model cells for up to eight distinct project-owned tests from history, with coverage, shared loss verdicts, fallback and outcome counts, and per-tool rates and Wilson ranges. Each test includes status; non-completed tests retain coverage only, with all tools unmeasured and null measurement… - [Get Agent Picks test history](https://docs.devtune.ai/api-reference/get-agent-picks-test-history.md): Read a page of the setup’s paid and self-run tests newest first (limit or pageSize, default and maximum 100), or the one test test_id names, whichever setup it belongs to. Rates for the current project tool use completed cell rollups and Wilson 90% likely ranges. Unfinished tests have no rates. Each… - [Get Agent Picks Actions and re-test results](https://docs.devtune.ai/api-reference/get-agent-picks-actions-and-re-test-results.md): Read the setup’s loss Actions, or the one actionId names whichever setup it belongs to, with creation evidence, baseline cell pick rates and Wilson 90% ranges, latest paid re-test status, credits and measured Outcomes results, whose ranges are at the confidence their method decided the verdict at: 9… - [Get paid Agent Picks test](https://docs.devtune.ai/api-reference/get-paid-agent-picks-test.md): Read the latest paid project test, or the one test_id names, its estimate in credits, reservation, settlement and refund in credit micros, and every run in it. Failed runs and runs cancelled before they start are not charged; a run cancelled after it starts pays its recorded spend. Unknown spend on… - [Get Agent Picks starter test](https://docs.devtune.ai/api-reference/get-agent-picks-starter-test.md): Returns the project’s latest Agent Picks starter test as it runs: its status (queued, running, completed, failed or cancelled), when it was created and completed, planned vs completed runs, and each run’s request, agent and model, status, recorded sandbox attempts and, once classified, its pick (a t… - [Start Agent Picks starter test](https://docs.devtune.ai/api-reference/start-agent-picks-starter-test.md): Start the project’s free starter test on its saved setup. Each confirmed account-owner work domain gets one; queueing uses it up, and a failed or cancelled starter test does not give it back. Starter tests cannot be cancelled. Repository inputs are not supported; the body is empty or names the setup… - [List Agent Picks setups](https://docs.devtune.ai/api-reference/list-agent-picks-setups.md): List the project’s Agent Picks setups with what each tests, oldest first, and the catalogue ids a save names. kind is job (the project’s own job), copy (its copy of a public category) or category (a setup saved on a public category itself). snapshot_digest is what saveAgentPicksJob and saveAgentPick… - [Save Agent Picks setup](https://docs.devtune.ai/api-reference/save-agent-picks-setup.md): Save a setup for a published public category. A project has one setup per category and can add as many as it needs. Without setup, this updates the project’s setup for category_id, or adds one when the project has none for it: the project’s own copy of DevTune’s setup for the category, with its requ… - [Clear Agent Picks setup](https://docs.devtune.ai/api-reference/clear-agent-picks-setup.md): Remove a setup. Its tests and results stay readable, its schedule stops, and its suggested Actions that nobody has accepted are removed. With several setups, pass setup; without it the call is refused with 400 agent_picks_setup_required. Removing the last setup leaves the project with none until one… - [Draft Agent Picks job](https://docs.devtune.ai/api-reference/draft-agent-picks-job.md): Draft a custom job from the project’s product: the requests agents are given (add, recommend, hosted), the building blocks that decide built_its_own and no_provider, and competitor suggestions, plus the job line when job is null. A draft is a proposal; nothing reaches a test until the job is saved.… - [Draft Agent Picks tool signals](https://docs.devtune.ai/api-reference/draft-agent-picks-tool-signals.md): Draft one tool’s signals for a custom job from package registries and its own site: the packages, config names, import patterns, config files, config patterns and domains that name it in a run. Pass other_tools, the job’s other tools, to have the draft propose which of them run inside one another; e… - [Save Agent Picks job](https://docs.devtune.ai/api-reference/save-agent-picks-job.md): Save a custom job as confirmed, with the setup that tests it. A project can have several custom jobs, each its own setup. Without setup, this updates the project’s custom job with the same job line (ignoring case and surrounding spaces), or adds a setup for a new job; a job without a source category… - [List Library items](https://docs.devtune.ai/api-reference/list-library-items.md): Lists the project's Library items: Current versions only, unless includeSuperseded is true. Content drafts are listed only when the account has content drafts and the credential also holds actions.read. Pass pagination.nextCursor back as cursor, with includeSuperseded unchanged, for the next page; a… - [Upload a text document to the Library](https://docs.devtune.ai/api-reference/upload-a-text-document-to-the-library.md): Uploads text to the Library as a new item. With replaceItemId it becomes the new Current version of that upload, and the old version stays in its version chain as Superseded. Library search, and agents when Available to agents is on, use the text once indexing finishes. The write acts as the member… - [Search Library items](https://docs.devtune.ai/api-reference/hybrid-search-library-items.md): Searches the project's Library items by title and text, ranking keyword and meaning matches together; without query it lists items as List Library items does. Returns Current versions only, unless includeSuperseded is true. Content drafts are included only when the account has content drafts and the… - [Get a Library item](https://docs.devtune.ai/api-reference/read-a-library-item.md): Returns one Library item with its full text as contentText and contentMarkdown, its provenance and its version history. A content draft is returned only when the account has content drafts and the credential also holds actions.read; otherwise it returns 404, as does an item outside this project. Req… - [Update Library grounding availability or version link](https://docs.devtune.ai/api-reference/update-library-grounding-availability-or-version-link.md): Turns Available to agents on or off for an item (groundingEnabled), or links a Superseded item to the later version that replaced it in the same version chain (successorId; null removes the link). Send at least one of them; sending the values the item already has changes nothing. The write acts as t… - [Get the Knowledge Profile and proposals](https://docs.devtune.ai/api-reference/read-the-knowledge-profile-and-proposal-queue.md): Returns the project's Company context (companyContext), its agent preferences in profile (Tone and Instructions for agents), and its proposed changes, newest first: Pending, Approved and Dismissed. profile is null until a preference is first saved. Pass pagination.nextCursor back as cursor for more… - [Update one Knowledge Profile field](https://docs.devtune.ai/api-reference/update-one-knowledge-profile-field.md): Saves a direct edit to one field: company_context (Company context), tone (Tone) or saved_instructions (Instructions for agents). The value replaces the field at once, without going through proposal review, and applies to future agent tasks; it does not rewrite earlier outputs. Sending the same valu… - [Review a Knowledge Profile proposal](https://docs.devtune.ai/api-reference/review-a-knowledge-profile-proposal.md): Decides a proposed Knowledge Profile change. approve writes the proposed value into its field, replacing the current value; dismiss leaves the profile unchanged; restore returns a Dismissed proposal to Pending. Repeating an approve or dismiss that already happened returns the proposal unchanged, so… - [List webhook subscriptions](https://docs.devtune.ai/api-reference/list-webhook-subscriptions.md): Returns active webhook subscriptions for the project. Requires the `webhooks.read` scope. - [Create a webhook subscription](https://docs.devtune.ai/api-reference/create-webhook-subscription.md): Registers a new webhook subscription. Returns the subscription including the signing secret (only shown once). Each call creates a subscription, so a retry after a lost response creates a second one: list the subscriptions and delete the extra. Requires the `webhooks.write` scope. - [Deactivate a webhook subscription](https://docs.devtune.ai/api-reference/delete-webhook-subscription.md): Deactivates a webhook subscription: it stops receiving events and leaves the list. Deactivating it again returns the same result. Requires the `webhooks.write` scope. - [Rotate a webhook signing secret](https://docs.devtune.ai/api-reference/rotate-webhook-signing-secret.md): Generates a new HMAC-SHA256 signing secret for the webhook subscription. The old secret is invalidated immediately. The new secret is only returned once in this response; if you lose the response, rotate again. Requires the `webhooks.write` scope. - [Get the agent setup draft](https://docs.devtune.ai/api-reference/get-the-agent-setup-draft.md): Reads the setup draft of the unclaimed agent registration the token belongs to. A claimed or expired registration has none (404 registration_unavailable). Takes an agent access token, not an API key. Takes no query parameters. Requires the `setup.draft` scope. - [Replace the agent setup draft](https://docs.devtune.ai/api-reference/replace-the-agent-setup-draft.md): Replaces the draft whole. It validates as the post-claim API would: the Agent Picks setup and job as saveAgentPicksSetup and saveAgentPicksJob accept them, with ids in the published catalogue. A refused draft is 400 invalid_draft naming the first problem. Takes an agent access token, not an API key.… - [Get the Agent Picks catalogue for a draft](https://docs.devtune.ai/api-reference/get-the-agent-picks-catalogue-for-a-draft.md): Lists the published categories with their active tools, the test apps agents start in and the agents with their models that a draft’s Agent Picks section may name. Which agents a plan includes is known only after the claim. Takes an agent access token, not an API key. Takes no query parameters. Requ… - [Get the AI platform catalogue for a draft](https://docs.devtune.ai/api-reference/get-the-ai-platform-catalogue-for-a-draft.md): Lists the AI platforms DevTune probes, in display order. Which ones a plan includes is known only after the claim. Takes an agent access token, not an API key. Takes no query parameters. Requires the `setup.draft` scope. ## OpenAPI Specs - [openapi](/openapi.json) This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.