quality-weapon
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use local development tools such as
git diff,git status, andfindto inventory changes and locate plan documents. These commands are restricted to the local repository context and are necessary for the skill's primary function of auditing code. - [DATA_EXFILTRATION]: Analysis of the skill's instructions and the provided Python script (
scripts/extract-plan-items.py) confirms no unauthorized network operations or access to sensitive system files (e.g., SSH keys, environmental secrets) are performed. All data processing is local to the repository. - [EXTERNAL_DOWNLOADS]: Research documentation included with the skill references well-known technology organizations and services such as Google, Vercel Labs, and Prisma. These links are used purely for referencing industry best practices and do not involve downloading or executing untrusted code.
- [PROMPT_INJECTION]: The skill establishes strong operational principles, specifically emphasizing 'evidence over opinion' and treating the plan document as a source of truth. It contains no instructions aimed at overriding agent safety filters or bypassing system-level constraints.
Audit Metadata