Action-control infrastructure for autonomous agent systems.
AgentVeil helps teams inspect what agents can do before deployment, gate risky actions before execution, and produce signed evidence after execution.
- AgentVeil SDK - Python SDK for posture checks, runtime gates, signed receipts, reputation signals, and verifiable execution evidence.
- Lurkr - local-only scanner for risky AI-agent capability surfaces before deployment.
- Paperclip AVP Plugin - trust and reputation tools for Paperclip agent teams.
- Capability scanning before deployment.
- Runtime gates before high-risk actions.
- Signed receipts after execution.
- Reputation-aware trust signals for agent delegation.
- Verifiable credentials, DIDs, and portable evidence where useful.
- Why AI Agent Reputation Needs Both Link Analysis and Flow-Based Gating - position paper on reputation and Sybil-resistance composition for open AI-agent networks.
pip install agentveilLurkr can be used as a local CLI and in GitHub Actions:
- Website: agentveil.dev
- Email: ob@agentveil.dev