The Agent Skills Directory

[SAFE]: The skill performs an automated audit of repository activity using standard Git and shell commands. It reads project files (such as plan records and instruction files) and generates a report within the repository's own directory structure. The use of node --check for syntax validation of JavaScript files before parsing is a secure practice. No credential exposure or unauthorized network communication was identified.- [INDIRECT_PROMPT_INJECTION]: The skill processes project-controlled files (Git logs, Markdown plans, and JavaScript event logs) to generate activity reports.
Ingestion points: The skill reads commit messages, file status, and the contents of files like CLAUDE.md, AGENTS.md, and board-events.js.
Boundary markers: The instructions explicitly mandate a "mechanical evidence pass" and forbid interpreting motives or diagnosing the reader, reducing the risk of the agent following instructions embedded in the analyzed data.
Capability inventory: The skill utilizes git, find, awk, and node for data collection and analysis.
Sanitization: It employs node --check to verify the syntax of JavaScript files before they are processed by awk, and uses specific field-based parsing logic to handle structured event data.

qstack-reflect