github
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- COMMAND_EXECUTION (SAFE): The skill defines standard usage of the official
ghCLI for repository management tasks.\n- PROMPT_INJECTION (SAFE): No evidence of behavioral overrides, jailbreak attempts, or instructions to ignore safety guidelines.\n- EXTERNAL_DOWNLOADS (SAFE): The skill does not perform any external downloads of scripts, binaries, or packages.\n- DATA_EXFILTRATION (SAFE): There are no patterns indicating the unauthorized collection or transmission of sensitive data or credentials.\n- INDIRECT_PROMPT_INJECTION (SAFE): Although the skill ingests content from GitHub repositories (such as issue descriptions or logs), which is an inherent risk surface, it does not demonstrate any vulnerable interpolation or automated execution of that content.
Audit Metadata