Skip to content

Security: humancoreai/Gategraph

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
v0.28.2_STABLE yes
v0.28.2_STABLE yes
older v0.15.x snapshots best-effort review only
older v0.14.x and below historical/reference only

Scope of Security Claims

GateGraph provides deterministic governance/enforcement checks for explicit task metadata. It does not verify semantic truthfulness of caller-provided fields and does not provide autonomous security classification.

Security claims are limited to behavior covered by evidence suites and documented invariants.

Reporting a Vulnerability

Report issues privately to the project maintainer before public disclosure. If no dedicated security email has been published for the repository yet, use the repository owner contact channel.

A useful report should include:

  • affected version
  • reproduction steps
  • expected vs. observed behavior
  • whether the issue affects governance, enforcement, audit, release integrity, or documentation claims

Out of Scope

  • requests for autonomous risk classification
  • feature requests unrelated to a security defect
  • attacks requiring a caller to intentionally provide false trusted metadata without an upstream trust failure
  • multi-node or distributed consensus concerns in single-node releases

Development Keyring Boundary

GateGraph may include a local development keyring constant for deterministic tests and local evidence runs. This is not a production secret and must not be treated as deployable signing material. Production use requires operator-managed key material outside the repository.

The development keyring exists to make evidence reproducible; it does not grant external authority, does not bypass enforcement, and does not remove token binding or revocation checks.

There aren't any published security advisories