plan-test-coverage
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run standard local development commands, specifically
npm test -- --coverageandnpx vitest run --coverage, to gather test data. These are routine operations for a QA-focused skill. - [INDIRECT_PROMPT_INJECTION]: The skill processes external codebase artifacts (source code, routes, and test files) to derive user stories and audit test quality. This creates a surface where instructions embedded in the audited code could potentially influence the agent's planning output.
- Ingestion points: Source code, route handlers, and test files are read to reconstruct user stories (
SKILL.md,references/methodology.md). - Boundary markers: Not present; the skill relies on the agent's interpretation of the codebase.
- Capability inventory: Local shell execution for running test suites and coverage reports (
SKILL.md). - Sanitization: None specified for the content read from the filesystem.
Audit Metadata