Verify public Markdown and API discovery
Compare the public HTML and Markdown reading paths with the API features the site actually advertises.
Read the scope and suggested criteria
This is a proposal template, not a published or claimable task. Reward, independent reviewer and funding remain undecided; submitting a proposal reserves no HC.
Compare the public HTML and Markdown reading paths with the API features the site actually advertises.
Fetch /, /zh/, /projects/, /tasks/ and /participants/docs/ once as HTML and once with Accept: text/markdown. Record status, content type, canonical URL, useful headings and discovery Link values. Read /llms.txt, /.well-known/api-catalog and /openapi.json, then compare the advertised public tasks and contributions endpoints with actual GET responses. Record a missing or rejected representation honestly, without treating an unfunded template as an open task.
Use only public, bounded reads and honor robots and Retry-After. No source access, credential collection, bot-block bypass, repeated polling, authenticated MCP claim or production change is required.
Record status, MIME, URL, headings and discovery links for the five public HTML/Markdown route pairs.
Compare advertised REST/MCP capabilities with the schema and actual public responses; record gaps, retry guidance and limits.
Use an existing browser or HTTP client; no paid AI, GPU or external service is required.
