Checks an agent’s success claims against its tool calls and the state it left behind.
-
Updated
Nov 6, 2026 - Python
Checks an agent’s success claims against its tool calls and the state it left behind.
A public slice of a solo operator's hash-chained record of AI-agent work being graded, refused and voided. Paper-only. Run verify.py in 60 seconds.
Evidence-first runtime for AI agents: record, verify and replay any action; catches false success (6.25% -> flagged).
Inoku Verified: an independent evaluation of how often an AI agent's 'done' is true. Method, benchmark and rules of the mark, published when ready to be checked.
ProofGate Runtime v0.2 - evidence-first claim verification (confirmed/refuted/unverified), HMAC-signed receipts, MCP server. Stdlib only; not independently audited.
Measure agent honesty across heterogeneous nodes: claim-vs-truth pass rate and false-success detection.
To associate your repository with the false-success topic, visit your repo's landing page and select "manage topics."