The Agent Skills Directory

[PROMPT_INJECTION]: The skill implements an ingestion workflow where external sources are copied verbatim into a raw/ directory and subsequently processed to generate wiki content. This exposes the agent to indirect prompt injection risks.
Ingestion points: Files located in raw/<domain>/<subdir>/ populated from external origin paths.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings for the content being processed from the raw/ directory.
Capability inventory: The skill utilizes file-read, file-write, and terminal execution (via pnpm) across the repository, including project metadata directories like .github/, .agents/, and .cursor/rules/ during refactoring operations.
Sanitization: No sanitization, escaping, or validation of external content is mentioned before the agent interpolates or reads the data to author wiki pages or perform linting.
[COMMAND_EXECUTION]: The skill requires the use of terminal commands to execute project-specific scripts (pnpm run check:wiki, pnpm run lint:md) and vendor-specific tooling (@paniolo/cli) for wiki validation and maintenance.
[EXTERNAL_DOWNLOADS]: The workflow relies on the Node.js ecosystem and specific vendor tools. These references to @paniolo/cli and pnpm are documented as part of the intended development infrastructure provided by the author.

paniolo-wiki