The Agent Skills Directory

[SAFE]: The skill provides a comprehensive health diagnostic suite for the OrchestKit ecosystem, focusing on manifest integrity, hook configuration, and environment validation.
[COMMAND_EXECUTION]: Employs shell scripts (check-mcp-pinning.sh, check-plugin-health.sh) and command-line utilities (jq, grep, wc) to audit the local project environment. These operations are limited to status reporting and validation of local files.
[REMOTE_CODE_EXECUTION]: Uses inline Python scripts for structured data processing (JSON parsing). These are implemented as heredocs within shell scripts and operate entirely on local input data.
[DATA_EXPOSURE]: Accesses project-specific files such as .mcp.json, .claude/settings.json, and .claude/memory/decisions.jsonl to report on configuration health and data integrity. No network exfiltration or unauthorized access patterns were identified.
[EXTERNAL_DOWNLOADS]: Mentions and checks for the presence of optional tools like agent-browser and tailscale. It provides recommendations for manual installation by the user rather than performing automated remote downloads.

doctor