test-harness

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions specify the use of 'git diff --name-only HEAD~5' during its reconnaissance phase to identify recently modified files. While this is a standard developer workflow, it involves executing local shell commands to gather metadata about the repository.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection (Category 8).
  • Ingestion points: The agent is instructed to read and analyze user-provided Python functions, classes, modules, and project-specific files like 'CLAUDE.md' or 'conftest.py'.
  • Boundary markers: The instructions do not define specific delimiters or security headers to separate untrusted source code content from the agent's instructions.
  • Capability inventory: The skill is capable of reading local files and writing new Python script files (e.g., 'tests/test_{module}.py').
  • Sanitization: There is no mention of sanitizing docstrings, comments, or configuration data which could contain malicious payloads designed to influence the generated code or subsequent agent actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 08:58 PM