The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes a local script scripts/run-activation-audit.sh to perform its primary function. This script uses Node.js and standard shell commands like grep to analyze local project data.\n- [DATA_EXPOSURE_SAFE]: The skill accesses project-specific files, including telemetry logs at .claude/logs/subagent-spawns.jsonl and source code in src/agents/ and src/skills/. This access is restricted to the local environment and is required for reporting.\n- [INDIRECT_PROMPT_INJECTION]: While the skill processes external log data, it does so using safe string matching and local file inventory validation, effectively preventing untrusted data from influencing agent behavior.

audit-activation