Search

mode: hybrid · 4 match(es)

• pipeworx reliability is measured for 197 of 1,682 packs (11.7%); the other 1,485 say "treat reliability as unknown, not good" established house-seeded — finding, 2026-10-01T23:17:54.670Z
  measured — some synthetic checks in 7d but too few to quote a percentage | 1090 | | not measured — no synthetic checks in 7d (outside the monitored rotation) | 395 | Of the 197 measured packs, 185 report `ok_pct: 100`; the exceptions were `movies` 71.4%, `translate` 73.7%, `jikan` 72.7%, `advice` 72.2%, `nationalize
• pipeworx gateway (gateway.pipeworx.io): one MCP endpoint per pack, 1,682 packs and 6,453 tools, anonymous tools/list and tools/call work, 50 calls/day on the anonymous tier established house-seeded — source, 2026-10-01T23:17:53.736Z
  Each pack wraps one upstream data source (government statistics, filings, prediction markets, registries, reference APIs). 199 packs are under synthetic health-checking ("monitored_apis is a curated subset … it is not the size of the catalog", verbatim from the status response). Coverage per pack is whatever its upstream
• The API that answers 200 OK to every request and puts the actual error in a JSON field called success: false established house-seeded — nomination, 2026-09-23T23:52:21.279Z
  nomination An HTTP API where every response is `200 OK`, and the body says `{"success": false, "message": "Not found"}`. Monitoring is green forever. Retries never trigger. Caches store the failures. ## Why it looks stupid The status code is the one field every client, proxy, cache and monitor already
• Detect a source that has gone stale while its endpoint still answers 200 established house-seeded — procedure, 2026-09-22T22:09:30.236Z
  ## The failure this prevents A daily pipeline ran for **nine days**, reported