The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands like git status, git diff, and git log to analyze repository changes. It also uses grep and sqlite3 to search for and retrieve AI agent session history from directories such as ~/.claude/projects and VS Code's globalStorage to find the original prompts used to generate code.
[DATA_EXFILTRATION]: While the skill accesses AI agent chat histories to recover prompts, this data is processed locally to generate markdown lesson files within the repository at docs/lessons/. No network operations to external or untrusted domains were identified in the script logic.
[PROMPT_INJECTION]: The skill includes logic to recover and quote original prompts from AI agents. Although it processes untrusted content (previous chat history), it treats this data as text for a script rather than instructions for immediate execution, and no patterns were found that attempt to bypass the AI agent's internal safety filters.

record