The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs automated validation of code examples and specifications using local system tools as part of its core workflow.
Evidence: Use of python -m doctest, pytest --doctest-modules, tsc --noEmit, and npx @redocly/cli lint for validating documentation accuracy.
[EXTERNAL_DOWNLOADS]: The documentation workflow recommends installing various standard linting and static site generation tools from official registries.
Evidence: Instructions to install pydocstyle, interrogate, linkchecker, eslint-plugin-jsdoc, and broken-link-checker via standard package managers.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied code and specifications, which it then executes or lints, creating a potential vector for indirect prompt injection if the user code is malicious.
Ingestion points: The skill reads user-provided source files (Python, TypeScript, JavaScript) and API specifications (OpenAPI, GraphQL) to generate and validate documentation.
Boundary markers: The skill workflow does not specify the use of boundary markers or instructions to ignore embedded commands when performing validation steps.
Capability inventory: The skill possesses the capability to execute code via python and pytest, and run shell commands via npx and local binary execution (tsc, linkchecker).
Sanitization: No explicit sanitization or filtering of code blocks within user-provided files is described before they are passed to validation tools.
[DYNAMIC_EXECUTION]: The skill utilizes automated testing frameworks to dynamically execute code snippets found within documentation strings.
Evidence: Automated execution of Python doctest blocks and pytest module-wide checks to ensure documentation examples are functional.

code-documenter