What problem does it solve? Monitoring NanoClaw agents, sessions, and messaging channels requires running CLI commands by hand and reading raw JSON output, with no unified visual overview of agent health, message activity, logs, or configuration files. ## Core Features & Use Cases - Runtime-derived dashboard: Point clidash at any CLI that lists resources as JSON (ncl, docker, kubectl) and it builds one tab per resource with a generic table, so new resources and columns appear with zero code changes. - NanoClaw-aware panels: Ships pre-wired with an agents overview (status cards joining groups, sessions, and wirings), message-activity charts read from session SQLite DBs, an allowlisted log tail, and a read-only viewer for group skills, CLAUDE.md, and profiles. - Read-only by construction: The server only executes configured argv templates via execFile with allowlist-validated resource substitution, never a shell, and binds 127.0.0.1 by default. - Use Case: After deploying NanoClaw, install clidash to watch all agent groups' container status, last-active times, and inbound/outbound message volumes from a single browser page at http://127.0.0.1:4690. ## Quick Start Ask the assistant to run the add-clidash skill to copy the clidash tool into tools/clidash, create the config from the example, and start the dashboard server.