The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes source code, docstrings, and configuration files as the source of truth for documentation updates. This creates a surface for indirect prompt injection where instructions embedded in the code could influence agent behavior.
Ingestion points: Source code, docstrings, pyproject.toml, and zensical.toml are read by the agent as instructed in SKILL.md.
Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands in the processed data.
Capability inventory: The skill can write files, execute a local Python script, and run shell commands like pytest, lint, and fmt tools.
Sanitization: Content is processed and rewritten without specific sanitization steps to filter out embedded instructions.
[COMMAND_EXECUTION]: The skill invokes several local commands and scripts to fulfill its documentation and validation tasks.
It runs scripts/gen-ref-pages.py to generate the project's reference documentation structure.
It executes validation tasks using uv run pytest for doctests and tools like rumdl, mise run lint, and mise run docs:build as specified in SKILL.md.
[DYNAMIC_EXECUTION]: The skill instructs the agent to write compact doctests in public APIs and then execute them using pytest for validation. This involves generating executable Python snippets within docstrings that are subsequently run in the development environment.
[EXTERNAL_DOWNLOADS]: The helper script gen-ref-pages.py specifies a dependency on the tomlkit package, which is automatically fetched from the Python Package Index (PyPI) via the uv tool at execution time. This follows standard Python dependency management practices.

rewrite-python-docs