Skip to content

Method

How the work on this site earns the word proven. One discipline runs through the catalog, the Agentic Runtime and RAWLY: AI does the work, gates decide what ships, and every claim traces back to a versioned artifact.

Prove it live, fail closed

A workflow is called proven here only after a real run traversed it end to end, with eval-gates enforced at every step and the per-step verdicts committed as verbatim JSON traces. A run that merely looks finished is not evidence — the trace is.

Fail-closed means a run has three honest outcomes, and all three are published:

Outcome What it means Exhibit
completed Every gate passed; the deliverable shipped Nine workflows on the live proofs page
returned A counter-review gate sent the work back; the final report was withheld by design WF-008
gate-halt A gate rejected its input and stopped the chain WF-009's second trace

The last two are not failures of the method — they are the method working. A returned audit and a halted chain prove the gates discriminate: they can say no, so their yes means something.

AI drafts, gates decide, a human signs off

Across projects, the same pattern repeats:

  1. Expert agents draft in shared context — domain agents chained by BPMN workflows in the runtime, a nutrition expert and a copywriter for RAWLY articles.
  2. An isolated checker verifies blind. It receives none of the drafting context and re-derives every claim from the source of authority — RAWLY's fact-checker re-resolves every PMID and DOI from scratch against the live registries, never from memory.
  3. A human gate closes the loop. Nothing counts as validated until it has been exercised live — hard refresh, real clicks, mobile and desktop, light and dark.

The middle step exists because generated citations can look perfect and be wrong: an audit of RAWLY's legacy articles exposed four hallucinated PMIDs, which is why the checker now runs isolated on every new piece. See the RAWLY content pipeline for the checker's design, and the runtime architecture for how gates are enforced in code.

Versioned, or it didn't happen

Every repository in this system carries the same regime: SemVer, annotated git tags, GitHub Releases, and a dated changelog entry per release.

  • The catalog versions its 38 agents and 37 skill folders as code, indexed in a CI-validated sidecar.
  • The runtime commits its live-run traces verbatim under docs/audit/live-runs/ — the same files the live proofs page quotes.
  • This site follows the identical discipline: each release folds its changelog entry, and the model that did the work is recorded per release.

Explore the evidence