The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the grep utility to search files within the <wiki-root>/wiki/ directory as a fallback when the catalog index is insufficient to answer a query.
[PROMPT_INJECTION]: The skill processes potentially untrusted data from the user's wiki and raw files to synthesize answers, which represents an indirect prompt injection surface. * Ingestion points: Reads information from index.md, various markdown files in the wiki/ directory, and source documents in the raw/ directory. * Boundary markers: The instructions do not provide explicit delimiters or ignore-instructions to help the agent distinguish between data and potential embedded commands. * Capability inventory: The agent possesses file system read access to the wiki root, write access for logging and page creation, and execution access for searching via grep. * Sanitization: The workflow does not specify any sanitization, validation, or escaping of the content ingested from the knowledge base.

wiki-query