The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell utilities including grep, awk, and git to perform validation of document states and manage file movements within the project structure.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local markdown files (_index.md, task_NN.md) to verify project status before execution.
Ingestion points: Reads status fields from task files and metadata rows from the reference index file.
Boundary markers: No explicit delimiters are used for the ingested file content.
Capability inventory: Includes directory creation (mkdir) and file movement (git mv) based on successful validation.
Sanitization: Employs an awk script that validates table headers, row formatting, and restricts paths to safe basenames to prevent directory traversal.

archive-spec