The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's core functionality involves ingesting and analyzing external code files and then running validation tools on them, which creates a vector for instructions embedded in data to influence the agent.
Ingestion points: Workflow steps for discovering, detecting, and analyzing undocumented code (SKILL.md).
Boundary markers: The instructions do not specify the use of delimiters or warning messages to separate user-provided code from instructions.
Capability inventory: The skill directs the agent to run python -m doctest, pytest, tsc, and npx tools on user-provided code (SKILL.md).
Sanitization: No sanitization or filtering logic is described for the content being processed before execution.
[COMMAND_EXECUTION]: The skill provides explicit instructions for the agent to execute several command-line development tools.
Evidence: Workflow validation steps include python -m doctest, pytest --doctest-modules, tsc --noEmit, and npx @redocly/cli lint (SKILL.md).
[EXTERNAL_DOWNLOADS]: The skill references external documentation and suggests installing packages from well-known registries.
Evidence: The skill links to its own documentation hosted on GitHub Pages (SKILL.md), which is a vendor-owned resource on a well-known service.
Evidence: Instructions recommend installing standard utility packages such as pydocstyle, interrogate, and eslint-plugin-jsdoc from official package registries (references/coverage-reports.md, references/documentation-systems.md).

code-documenter