cleanup-feature-flags
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a structured safety framework that classifies the readiness of a feature flag for removal. It explicitly blocks operations if environments are inconsistent, if experiments are active, or if dynamic keys are detected, ensuring no accidental breakage.
- [COMMAND_EXECUTION]: The skill performs shell commands to run the repository's own build, lint, and test suites (SKILL.md, Phase 6). This is a standard and safe practice for validating code modifications in a development environment.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external content from the application repository and feature flag definitions (SKILL.md, Phase 4). Boundary markers are implicitly defined through the use of specific SDK patterns for code search. The skill's capabilities include executing local tests and opening pull requests. Risks are mitigated by the prescriptive logic used to resolve flag treatments and the requirement for explicit human approval before any code is pushed or a PR is opened.
Audit Metadata