Skip to content

Security: iriseye931-ai/agentcodehandoff

Security

SECURITY.md

Security Policy

Scope

AgentCodeHandoff is a local-first terminal tool. Most risk comes from:

  • local command execution through installed agent CLIs
  • filesystem access inside repos you point it at
  • background bridge processes on the local machine

It is designed around a bring-your-own-agent model:

  • users install and authenticate their own local agent CLIs
  • AgentCodeHandoff coordinates those local tools
  • AgentCodeHandoff is not intended to proxy or resell third-party subscription credentials through a hosted service

Reporting

If you find a security issue, do not open a public issue first.

Report it privately to the maintainer and include:

  • affected version or commit
  • impact summary
  • reproduction steps
  • whether it requires local access, repo access, or a malicious agent/tool response

Current Security Expectations

  • review bridge automation before enabling it in sensitive repos
  • prefer least-privilege local environments
  • treat agent CLI auth state as sensitive local runtime context
  • validate repo and environment assumptions before using unattended bridge automation in production workflows
  • prefer local user-controlled runtimes over shared credential relays

There aren't any published security advisories