The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a bundled Python script, scripts/lint_wiki.py, to perform integrity checks, such as identifying broken links, orphan pages, and schema violations. This script operates locally and is limited to scanning the markdown files within the user-configured wiki directory.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it is designed to ingest and process untrusted data from external sources (e.g., web articles and PDF papers) to populate the wiki. Malicious instructions embedded in these sources could potentially influence the agent's behavior during the ingestion or summarization phase.
Ingestion points: External data is fetched via tools like web_extract and stored in the raw/ directory of the wiki.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent reads from the raw/ source files.
Capability inventory: The skill leverages file management capabilities (read_file, write_file, patch, search_files) and local script execution to maintain the knowledge base.
Sanitization: There is no evidence of explicit sanitization or filtering of the content extracted from external sources before it is processed into permanent wiki pages.

md-wiki