A202 MCP Server
About
Official reference MCP server for A202, the Verifiable Agreement Protocol for Agent-Led Commerce: signed mandates, agreement formation, obligations, evidence verification, and hash-chained transaction records for direct agent-to-agent B2B commerce.
Details
- Author
- a202-protocol
- Categories
- Finance
Jump to
Setup
Install A202 MCP Server in your MCP client (Claude Desktop, Cursor, Windsurf, and others).
Repository: https://github.com/a202-protocol/a202
Follow the installation instructions in the repository README, then restart your MCP client.
A202: the Verifiable Agreement Protocol for Agent-Led Commerce
A202 is verifiable commerce for agent-to-agent or agent-led transactions: an open specification of commercial authority, negotiation state, and verifiable conformance for transactions between independent organisations, including transactions conducted on their behalf by software agents.
It defines typed objects for delegated commercial authority, a state machine for the transaction and for each bilateral session inside it, rules for what may be disclosed to whom, and an executable conformance suite that turns each of those into a check an implementation either passes or fails.
The full statement of purpose, scope, and non-goals is inCHARTER.md.
Created and developed by A. A. Musse. SeeMAINTAINERS.md.
- v0.1.0is the first tagged release of the set: a tag, a digest for every schema file, the conformance manifest, and release notes, published together. SeeRELEASES.mdandCHANGELOG.md. Before 1.0 a MINOR increment may break compatibility, and any break carries migration notes.
- The name isA202, spoken "A two-oh-two", and in fullA202, the Verifiable Agreement Protocol for Agent-Led Commerce. The long form is a descriptor and not an expansion: the letters do not stand for it. The202is HTTP 202 Accepted, whichA202-0017makes the status an accepted submission returns, because acceptance is the primitive the rest of the specification is built on. TheA202-reason-code prefix, theA202-NNNNproposal identifiers, and thea202-commercial/0.1specification version string all follow from the name.
- A202™ is a trademark of Plural Worlds. Permitted use of the name is stated inTRADEMARK.md.
- Schema$idvalues resolve underhttps://schemas.a202.org. Fixture hosts use reserved.invalidnames, because test data must never resolve.
- Licensed under the Apache License, Version 2.0.One licence covers the whole repository: specification text, schemas, fixtures, manifest, runner, and informative documents. The licence carries an express patent grant from each contributor. SeeLICENSEandCONTRIBUTING.md.
- External contributions are accepted on the terms inCONTRIBUTING.md: inbound contributions under the same licence, with a developer certificate of origin sign-off.
Each specification document carries a status header stating which of its sections are normative and which are informative.
The runner validates every fixture named in the manifest against the schemas, then applies the invariants that JSON Schema cannot express. Schema validity is not conformance, which is the reason the runner exists.
It needsjsonschema>=4.18. If that is not on the system interpreter, a virtual environment is enough:
python3 -m venv .venv && .venv/bin/pip install "jsonschema>=4.18"
python3 conformance/run-conformance.py --verbose
The expected result is every fixture passing and none failing, with the totals the manifest carries: the manifest is the single source for the count, and the runner prints it on every run. The runner also asserts that each negative fixture is refused for the reason code the manifest declares for it, wherever the normative layer raises codes at all. Run it before and after any schema change.
Every negative fixture is minimal: removing the single offending element must leave a document that validates cleanly. A negative fixture that fails for an incidental reason tests nothing, so verify that when adding one.
The suite does not depend on anyone remembering to run it. It runs, together with the reference implementation tests and the MCP server tests, on every pull request and on every push to the default branch, under.github/workflows/checks.yml.GOVERNANCE.mdsection 3.4 requires the suite to pass for any change to schemas, fixtures, the manifest, or the runner, and that workflow is what turns the requirement into a gate.
- CHARTER.mdfor what this is and what it deliberately is not.
- schemas/canonical-commercial-model-v0.1.mdfor the object model, the envelope, and the invariants schema validation cannot express.
- negotiation/pilot-transaction-state-machine-v0.1.mdfor what moves state and what does not.
- conformance/manifest-v0.1.jsonfor the fixtures that decide whether an implementation agrees with either of the above.
Bridge Town is an MCP-native, git-versioned financial modeling platform for FP&A teams and finance leaders. AI agents use Bridge Town tools to create projects, write Python model files, run models in isolated cloud sandboxes, query data, write outputs to Google Sheets, create dashboards, branch scenarios, and collaborate with teammates.
The Capital.com MCP Server lets your AI assistant talk to your trading account directly. Market data, position checks, trade previews – all in plain language, without leaving your AI tool.
Coinrule Agentic Trading MCP enables investors to create, backtest, execute, and manage trading agents through natural language across stocks, crypto and ETFs
Invest with Claude and other AI assistants
Australian Consumer Data Right Product Data
Remote MCP server for historical crypto & prediction-market data: search ~500K instruments, live market stats (OHLC, turnover, spreads, depth, slippage) and tick-data purchase. Keyless for catalog & stats; optional OAuth for account tools. Endpoint: https://cryptostruct.com/mcp
Cross-border debt collection from your AI assistant: check cases, get pricing, submit new cases.
Read-only MCP server for your Evibe investment portfolio + live market data (holdings, performance, dividends, benchmarks, screeners). Works with Claude & ChatGPT.
Financial and quantitative modeling engine for AI agents. Typed, named, deterministic.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
