Skip to content

Security: mentu-ai/commitment-protocol

SECURITY.md

Security Policy

Reporting a vulnerability

Do not open a public issue for a security problem.

Report it through GitHub's private vulnerability reporting, with a description, the steps or input that reproduce it, its impact and, if you have one, a suggested fix. We acknowledge within 48 hours and give a timeline.

Scope

This repository is a specification and a reference verifier. In scope:

  • False verification: any input on which tools/verify_ledger.py returns verified for a ledger that was altered in a way the specification says a verifier detects, or returns verified where it should return incomplete.
  • Canonicalization mismatches: any signal for which the specified hash algorithm and the reference verifier disagree.
  • Specification gaps with security consequences: wording that permits an implementation to report closure, verification or authority that the records do not support.

Out of scope: vulnerabilities in a particular implementation of the protocol. Report those to that implementation's maintainers.

Known limits

The content hash excludes prevHash, and no verification profile detects a rewrite that keeps every row placed on the chain; detecting one needs a construction binding each row to its ancestry, which v2.3 does not define. LEDGER.md §Verdict and coverage states this. A report that only restates it is not a vulnerability; a verifier that returns verified on a ledger where the specification says it must not is.

There aren't any published security advisories