The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and refactor local source code and test suites, creating a surface where malicious instructions embedded in the repository code could influence the agent's behavior during a review or adversarial sweep.
Ingestion points: Local source code files, test files, and contract documentation (.md) read by the agent or scanner script (e.g., scripts/analyze_maintainable_tests.py).
Boundary markers: Absent. The skill instructions do not explicitly require the use of delimiters or system-level instructions to ignore embedded prompts in analyzed code.
Capability inventory: The agent possesses file reading, file writing (for repairs and new tests), and command execution (running the scanner script) capabilities.
Sanitization: Absent. The skill does not describe any specific sanitization or validation of the content of the files it processes.
[COMMAND_EXECUTION]: The skill includes a Python-based scanner utility and instructions for its execution (python3 scripts/analyze_maintainable_tests.py <path>). Additionally, the accompanying test suite (scripts/test_analyze_maintainable_tests.py) uses subprocess.run() to execute this scanner during verification. These commands are local, utilize relative paths to the skill's own scripts, and serve the primary functional purpose of the skill.

maintainable-tests