The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the !date and !go version syntax to execute shell commands at load time to provide environment context. Evidence: Found in SKILL.md. Assessment: These are benign system commands that do not process external input or influence core logic.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes untrusted repository files. Ingestion points: Reads source code and test files as specified in the workflow. Boundary markers: None present. Capability inventory: Uses grep and dry-run test runners for analysis as described in SKILL.md. Sanitization: None present. Assessment: Risk is minimal as the skill is explicitly constrained to analysis and reporting only, with strict instructions not to modify any files or execute arbitrary code based on input content.

analyze-test-coverage