# Agentic Keychain

> Autonomous trading agents on Base and Solana: official stand per live lane, closed trades only, self-reported, signed and hash-chained. No capital accepted.

Keychain Trading is the trading side of Agentic Keychain and a development preview: the trading system computes its own results (self-reported); each published snapshot is signed and hash-chained, which shows it is unchanged since publication, not that the trades were executed. There is no trading API, wallet connection, order route or deposit address.

Next to it: Agentic Keychain (Agent Skill ROI Registry & Benchmarking) is a registry that measures reusable agent skills and sells some of them, including its own, through Polar as merchant of record. If you are deciding whether a skill could make a task cheaper, resolve_task gives one compact answer for a task in plain words, and evaluate_capability returns the same deterministic recommend / not recommended / insufficient evidence answer with the arithmetic; not recommended is a normal answer. Ranking never depends on payment, and publisher claims never decide. These tools never charge; any purchase needs your operator to complete a hosted checkout.

This is a commercial registry. It sells some of the skills listed here, through Polar as merchant of record, and its operator earns money when one is bought. The evaluate answer can be "do not buy": it says so when the measured evidence does not cover your task, when the price is not recovered, or when a free alternative is enough. Every purchase ends in a checkout that a human completes.

Fields this API version does not define can appear and have no meaning for a client that does not know them. A kind, delivery or pricing.model value that a client does not know means the capability is not supported by that client; it is never purchasable by inference.

## Keychain Trading

- [Keychain Trading](https://agentickeychain.com/trading/index.md): development preview; self-reported, signed results of autonomous trading agents, the official stand per live lane, read-only as JSON at https://agentickeychain.com/trading/results.json, https://agentickeychain.com/trading/history.json and https://agentickeychain.com/trading/proof.json; methodology at https://agentickeychain.com/trading/methodology/index.md
- [Keychain Trading methodology](https://agentickeychain.com/trading/methodology/index.md): how a trade is counted, how each figure is computed and checked, and its limitations

## Use

- [For agents](https://agentickeychain.com/agents/index.md): connect over MCP (https://agentickeychain.com/mcp) or HTTP, evaluate, buy with your operator's approval
- [Method](https://agentickeychain.com/method/index.md): evidence labels, statistics, sale and ranking rules
- [OpenAPI](https://agentickeychain.com/openapi.json): the HTTP API of the skill registry
- [All skills](https://agentickeychain.com/skills/index.md): every skill with its state, price, evidence label and decision; the same list as JSON at https://agentickeychain.com/skills/index.json
- [Tasks](https://agentickeychain.com/capabilities/index.md): the agent tasks the registry covers and the skills listed for each
- [Agent Skill](https://agentickeychain.com/.well-known/agent-skills/agentic-keychain/SKILL.md): the free access skill for this registry; the Claude Code plugin marketplace is https://agentickeychain.com/integrations/claude-code/marketplace.json

## Skills

- [Log Triage](https://agentickeychain.com/skills/log-triage/index.md): Streams a large log once and prints a bounded Markdown summary - levels, first error, grouped messages, stack traces, time span.
- [Repo Cartographer](https://agentickeychain.com/skills/repo-cartographer/index.md): One-command Markdown map of a repository (languages, layout, manifests, entry points, symbols) sized to a character budget.

## Tasks

- [Explore an unfamiliar code repository](https://agentickeychain.com/capabilities/explore-repository/index.md): Build a working picture of an unfamiliar repository: languages, layout, entry points, build files and where the main code lives, before reading single files.
- [Find where something is implemented](https://agentickeychain.com/capabilities/find-code/index.md): Locate the files, classes and functions that implement a feature or symbol in a repository without reading the whole tree.
- [Triage a large log file](https://agentickeychain.com/capabilities/triage-logs/index.md): Summarize a log file too large to read into an agent's context window: error levels, the first failure, recurring messages, stack traces and the time span.

## Optional

- [Agent manifest](https://agentickeychain.com/.well-known/agentic-keychain.json): the entry points, the evidence labels and where the JSON form of each page lives
- [Everything in one file](https://agentickeychain.com/llms-full.txt)
- [Sitemap](https://agentickeychain.com/sitemap.xml): every indexable page with its last content change
- [Agent Skills discovery index](https://agentickeychain.com/.well-known/agent-skills/index.json): the free access skill in the Agent Skills discovery format
- MCP server name: `com.agentickeychain/agentic-keychain`, the name in the official MCP registry's server.json
- [Signing keys](https://agentickeychain.com/keys.json)
- [Current release](https://agentickeychain.com/release.json)
- [Transparency log leaves](https://agentickeychain.com/log/leaves.json): the leaf hash of every logged entry (RFC 9162 Merkle tree, SHA-256)
- [Signed tree head](https://agentickeychain.com/log/sth.json): the log's current size and root, signed with a key in keys.json