Codex Security deep scan
Plans repeated, multi-pass security scans across an authorized repository scope.
OpenAI plugin examples
What this skill does
Use for explicitly requested exhaustive or variance-reducing repository/security scans.
Where, why and how to use it
Use only for a deep scan request. Define the authorized inventory, multiple independent passes, evidence tracking, and a final reconciliation.
This differs from a standard scan and can consume substantial time/resources. Source file existence does not establish any scan result.
Source location
openai/plugins:plugins/codex-security/skills/deep-security-scan/SKILL.md
Compatibility and setup
Codex Security host workflow; not portable as a simple standalone prompt.
Requires scoped code access, sufficient runtime budget, and the security plugin's workflow context.
Boundaries
No production testing or boundary expansion; keep evidence free of secrets and sensitive user data.
Source version
openai/plugins main; unpinned. Reference checked 2026-10-14. Source verification does not mean this skill is installed, its connector is available or its service entitlement active.
Example instruction
Do a multi-pass security scan of this authorized repository, with separate coverage and validation evidence.