| Version | Supported |
|---|---|
| 0.1.x | Yes |
Do not open a public GitHub issue for security vulnerabilities.
To report a security issue, email the maintainers directly or use GitHub's private security advisory feature:
GitHub advisory: https://github.com/WeepsDanky/agentpod/security/advisories/new
Please include:
- A description of the vulnerability
- Steps to reproduce
- Potential impact
- Any suggested mitigations
We will respond within 5 business days and aim to release a patch within 30 days for confirmed issues.
Issues in scope:
- Authentication or token bypass in the hub join/exchange/renew flow
- Context export boundary violations (data leaving the machine that shouldn't)
- Privilege escalation via policy guard logic
- Injection vulnerabilities in task input handling
Out of scope:
- Vulnerabilities in dependencies (report upstream)
- Issues only reproducible with a maliciously modified hub
- Theoretical attacks without a practical reproduction path
AgentPod is local-first. Each agent remains responsible for its own boundaries.
Key security guarantees in v0.1:
- Only explicit
payloadandattachmentsfields leave the machine on delegation - Full conversation transcripts, hidden prompts, and implicit memory never leave by default
- The task deduplication registry prevents the same
task_idfrom executing twice - Bearer tokens are short-lived and bound to
peer_idandkey_fingerprint
See docs/protocol-v0.1.md for the full context export boundary definition.