The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes external test files and directories, creating an indirect prompt injection surface where instructions hidden in the code being reviewed could target the agent.
Ingestion points: The skill ingests untrusted data from the path provided in the <test_file_or_directory> argument as defined in SKILL.md.
Boundary markers: No specific boundary markers or delimiters are defined in SKILL.md to isolate external data from the prompt context.
Capability inventory: SKILL.md instructs the agent to perform command execution via pytest and use the python-engineering:python-pytest-architect tool.
Sanitization: The skill does not implement specific sanitization or filtering of the ingested content.
[DYNAMIC_EXECUTION]: The skill requires the execution of code through pytest and potentially other tools to validate test behavior and collect execution evidence. This dynamic execution is restricted to the local environment and is necessary for the skill's primary purpose of test auditing.

comprehensive-test-review