threads
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed with extensive security guardrails to manage autonomous multi-agent repository workflows without introducing malicious patterns.
- [DATA_EXPOSURE]: Implements a robust 'forbidden_files' policy that explicitly prevents agents from accessing or modifying sensitive configuration files such as
AGENTS.md,CLAUDE.md, and environment settings. - [INDIRECT_PROMPT_INJECTION]: While the skill ingests external data from GitHub issues and PRs, it mitigates potential injection risks by incorporating a mandatory 'Reviewer' lane tasked with identifying 'security and injection risks' before any code implementation is accepted or merged.
- [COMMAND_EXECUTION]: Uses standard development CLI tools (
git,gh) within isolated worktrees to ensure that parallel tasks do not conflict and that all operations are traceable to specific PRs and branches.
Audit Metadata