security-audit-skill

A coding-agent skill for multi-phase security audits with independently verified, machine-readable findings

More in Security

SDKs and CLI for Codex Security

Security

Dogwood is a policy language for authorization decisions that depend on history — not just a single request, but patterns of events over time.

Security

🔐 Egress credential injection for agents: agents hold placeholders, the cell fills real keys at the edge for approved hosts.

Security