test-harness
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands, specifically
git diff --name-only HEAD~5, to identify recently modified files and determine the scope for test generation. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted source code and project configuration files to generate test logic. It lacks explicit boundary markers or instructions to ignore embedded commands within the analyzed code, which could allow malicious comments or strings in the target files to influence the agent's behavior.
- Ingestion points: Source code files, function signatures,
CLAUDE.md,conftest.py, and project configuration files likepyproject.toml(SKILL.md, Phase 1). - Boundary markers: No delimiters or safety instructions are defined to separate untrusted code content from the generation prompt.
- Capability inventory: The skill generates executable Python code and performs file system reads and local command execution.
- Sanitization: There is no evidence of sanitization or filtering applied to the content of the analyzed files before they are processed by the LLM.
Audit Metadata