Evidence and limits
Evidence Note: Seven-stage agentic workflow (classify through report) with hard failure bounds and an approval gate before irreversible actions, plus a formalized fable-judge role that re-runs claimed checks by diffing and executing rather than reading reports; README cites 15 eval rounds across 260+ agent runs comparing bare models to method-guided models (e.g. Haiku catching 5/5 planted frauds with the method vs 3-4/5 without).
Last reviewed: 2026-08-29
What was reviewed: README reviewed only; no runtime test recorded
Source and review details
Snapshot Date: 2026-08-29
Snapshot Commit: 88b5cf36b10ee3679e08ee0f0181b9774d481508