The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from pull request diffs and documentation files.
Ingestion points: Capability source code under src/pydantic_ai_harness/ and documentation files in docs/harness/ (SKILL.md).
Boundary markers: The instructions do not define boundary markers or delimiters to separate instructions from the data being reviewed.
Capability inventory: The agent is authorized to run shell commands (uv run pytest) and execute deterministic code snippets.
Sanitization: The skill lacks explicit sanitization or validation of documentation snippets before they are executed.
[COMMAND_EXECUTION]: The skill instructs the agent to run uv run pytest tests/harness/test_doc_snippets.py to verify documentation parsing and harness imports (SKILL.md).
[DYNAMIC_EXECUTION]: The agent is required to 'Execute every changed deterministic snippet unchanged' to verify functionality, which involves runtime execution of code embedded in documentation files (SKILL.md).
[EXTERNAL_DOWNLOADS]: The skill references and links to official Pydantic resources at pydantic.dev and github.com/pydantic/pydantic-ai for verification purposes. These are recognized vendor assets.

docs-parity-reviewer