implement-issue
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: Implements comprehensive protections for sensitive information. The skill mandates redaction of HTTP(S) user credentials from URLs and git remotes before processing. It also enforces structured argument passing to CLI tools to prevent command injection from maliciously crafted repository names or issue URLs.
- [PROMPT_INJECTION]: Contains explicit architectural safeguards against indirect prompt injection. The core instructions define external data (issue bodies, comments, linked pages) as untrusted evidence rather than instructions, requiring the agent to ignore any embedded commands or attempts to override system behavior.
- [COMMAND_EXECUTION]: Uses industry-standard tools (gh, glab, git) in a constrained manner. The skill uses a "fail-closed" approach, requiring explicit authentication and verification of repository identity before any state-changing operations are performed.
- [EXTERNAL_DOWNLOADS]: Interacts only with well-known hosting services (GitHub, GitLab) using their official CLI tools. It explicitly forbids the automatic installation of software or authentication of accounts, ensuring the environment remains under user control.
Audit Metadata