Skip to content

Security: gexiro-global/ai-vault-contract

SECURITY.md

Security Policy

Reporting

Use GitHub private vulnerability reporting if enabled on this repository, otherwise email admin@gexiro.com.

The most useful report here is a case where the linter misses a secret-shaped string it should catch, or where the contract's OPSEC guidance is wrong or incomplete. Never include a real secret in a report — describe its shape.

The OPSEC rule this repo exists to protect

A versioned, mirrored vault keeps every write forever. The contract (AGENTS.md, OPSEC section) is blunt about it: secrets never go into a note, not even redacted, because redaction does not remove them from git history or any mirror. If one lands there, the primary control is to rotate the secret out-of-band immediately, then scrub history — in that order.

tools/lint_vault.py includes a coarse secret-shaped-string check as a backstop. It is a safety net, not a guarantee: pair it with a real pre-commit secret scanner (gitleaks or trufflehog) as the contract recommends.

There aren't any published security advisories