The Agent Skills Directory

[SAFE]: The skill implements a narrow retrieval architecture that limits the agent's file system interaction to specific, indexed paths, preventing unintended data exposure from wide directory scanning.
[SAFE]: The diagnostic script (scripts/smoke-system-qa.py) provides a safe method for verifying skill installation and configuration using standard Python libraries without external network dependencies or unsafe code execution.
[SAFE]: Instructional constraints, such as banning 'blind full-repo scanning' and requiring specific activation keywords, reinforce the agent's operational boundaries and prevent accidental safety filter bypasses.

fayanruju