froq-os-contrib
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements robust security boundaries by prohibiting the agent from filing PRs, commenting, or interacting with upstream repositories directly.
- [INDIRECT_PROMPT_INJECTION]: While the skill involves processing external data from GitHub issues, it effectively mitigates injection risks by limiting the agent's output to drafts and requiring the user to manually perform all upstream actions via a web interface.
- [COMMAND_EXECUTION]: Usage of tools such as
moandmiseis restricted to local repository management within designated paths, with no evidence of arbitrary or unsafe command execution patterns. - [DATA_EXFILTRATION]: No exfiltration vectors were identified; the skill contains explicit instructions to avoid using institution-specific resources for open-source work and strictly manages how identity is represented in commits.
Audit Metadata