The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's core workflow includes a validation step that executes local shell commands such as python -m doctest, pytest, and tsc --noEmit. These commands are used solely to verify the functional correctness of code examples and type definitions within the documentation.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx @redocly/cli to perform linting on OpenAPI/Swagger specifications. This command may fetch the latest version of the Redocly CLI from the public npm registry. Redocly is a well-known service for API documentation, and its CLI is a standard industry tool.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because its primary function involves reading and analyzing external source code provided by the user. If this code contains maliciously formatted comments or instructions, it could attempt to influence the agent's behavior during the documentation or validation phase.
Ingestion points: The skill reads and documents user-provided files during the 'Discover' and 'Analyze' phases of its workflow.
Boundary markers: There are no specific delimiters or "ignore instructions" warnings implemented to isolate the processed code from the agent's internal logic.
Capability inventory: The skill has the capability to execute shell commands (python, pytest, tsc, npx) through its validation logic defined in SKILL.md.
Sanitization: The skill does not perform specific sanitization or filtering of the ingested code content before it is processed or passed to validation tools.

code-documenter