The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands including grep and awk to search through and update the .knowhow/ workspace directory. These commands facilitate automated indexing, cross-referencing, and status signaling for knowledge management tasks.
[PROMPT_INJECTION]: The skill processes untrusted data from inbox files to generate structured content and search queries, creating an indirect prompt injection surface. While the risk is localized to the knowledge repository, the lack of sanitization on extracted keywords is noted.
Ingestion points: Reads content from all files within the .knowhow/inbox/ directory.
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded instructions within processed notes.
Capability inventory: Executes grep for searching and awk for updating signal files, and possesses write access to the .knowhow/ knowledge structure.
Sanitization: The skill does not implement explicit escaping or validation for strings extracted from inbox items before they are interpolated into shell commands.

knowhow-distill