fix
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill orchestrates the execution of standard development tools such as npm, bun, pytest, and the GitHub (gh) CLI. These tools are used for their intended purposes: running tests, performing type checks, and viewing CI/CD logs.
- [DATA_EXPOSURE_AND_EXFILTRATION]: No evidence of unauthorized data access or exfiltration was found. The skill's primary function involves reading project files and logs to diagnose bugs, which is consistent with its stated purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill inherently possesses an attack surface for indirect prompt injection as it ingests untrusted data from error logs and CI/CD outputs. However, the risk is mitigated by the structured nature of the workflows, which include human-in-the-loop review modes and mandatory verification steps (tests, linting) before finalization.
- [REMOTE_CODE_EXECUTION]: No patterns of downloading and executing remote code from untrusted sources were detected. The skill utilizes local Python scripts for internal tasks like classification and reporting, which do not perform any network operations.
Audit Metadata