For agents and developers
Read-only API
The services, method, evaluation standard, facts and published records on this site, as JSON. No personal data, no drafts, no write endpoints. Responses are cached for five minutes and open to cross-origin requests.
Endpoints
| Endpoint | Returns |
|---|---|
| GET /api/v1/services | The five services with purpose, buyers, deliverables, duration and boundary |
| GET /api/v1/services/{slug} | One service |
| GET /api/v1/method | The five delivery stages and the four autonomy stages |
| GET /api/v1/evaluation-standard | The seven measures and the reporting rule |
| GET /api/v1/facts | The facts page as an array with dates |
| GET /api/v1/stories | Published client stories only |
| GET /api/v1/benchmarks | The benchmark agenda; results only for released benchmarks |
| GET /api/v1/publications | Published publications |
| GET /api/openapi.json | OpenAPI 3.1 description of this API |
Downloads
Buyer scorecard (CSV)Five weighted criteria for comparing AI agent suppliers/downloads/buyer-scorecard.csvBuyer scorecard (JSON)The same, as JSON/downloads/buyer-scorecard.jsonBuyer scorecard (Excel)One-tab workbook with weighted totals/downloads/buyer-scorecard.xlsxFailure tests (CSV)Failure, control and test for each of the seven failure tests/downloads/failure-tests.csvFailure tests (JSON)The same, as JSON/downloads/failure-tests.jsonEvaluation standard (JSON)Measure, definition, unit and how it is reported/downloads/evaluation-standard.json
MCP server
A Model Context Protocol server exposes the same read-only data as tools: get_services, get_method, get_evaluation_standard, get_facts, list_client_stories, list_benchmarks and run_scorecard. Endpoint: https://mcp.chainfore.com (streamable HTTP).
Plain text and Markdown
Every page has a Markdown twin at the same path plus .md, and a request with Accept: text/markdown returns it. /llms.txt is the index; /llms-full.txt is every page in one document.