penholder
ai.penholder/penholder
80
security score
Preventive human-approval write-gate for AI agents: writes commit only after a human approves.
Healthy
Security scan
Last scanned 3 hours ago — static analysis of the published manifest.
- lowManifest makes network requests.
- mediumNo source repository linked — code cannot be audited.