testing-quality-gate
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has an inherent attack surface because it is designed to process repository content and execute project-defined scripts (build, lint, test).
- Ingestion points: The skill reads repository code, configuration files, and documentation to determine test boundaries and execution commands.
- Boundary markers: No explicit boundary markers or "ignore instructions" delimiters are defined in the skill instructions.
- Capability inventory: The skill allows the agent to execute any shell commands defined as repository targets for building, linting, and testing (documented in
SKILL.md). - Sanitization: No sanitization or validation of the repository-defined commands is specified beyond checking they are the "documented commands."
- [COMMAND_EXECUTION]: The skill instructs the agent to run project-wide checks, including build commands, linters, and test suites. This is consistent with the skill's primary purpose of serving as a quality gate for software development.
Audit Metadata