apply-qa-fixes
Pass
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill inherently possesses an indirect prompt injection surface because its primary function is to interpret and act upon natural language recommendations from external quality gate files. However, this is expected behavior for its stated purpose.
- Ingestion points: Quality gate YAML files in
.claude/quality/gates/and assessment reports in.claude/quality/assessments/. - Boundary markers: None identified; the skill is designed to trust the recommendations generated by its companion quality tools.
- Capability inventory: Includes reading, writing, and editing local project files, as well as executing shell-based testing and linting commands.
- Sanitization: No explicit sanitization of recommendation strings is performed, relying on the integrity of the trusted quality gate generation process.
- [COMMAND_EXECUTION]: The skill uses subprocess calls to execute local validation scripts and standard developer tools such as
npm run lintandpython .claude/skills/bmad-commands/scripts/run_tests.py. These executions are limited to the local development environment for verification and configuration reading.
Audit Metadata