setup

Configure this machine to report usage. Arguments, possibly empty: $ARGUMENTS

1. Check current config: cat ~/.usage-analyzer.json (missing file means not set up yet).
2. You need three things. Take them from the arguments, else from what the user already said, else ask for the missing ones in a single question:
   • name — how they appear on the dashboard (alice, Bekzod). Short, no spaces.
   • url — the dashboard URL the team already runs, e.g. https://usage.example.com. If nobody runs one yet, say so and point at the "Server" section of the plugin README; do not start a server on their laptop unless they ask for it.
   • token — the shared secret that dashboard was started with. Optional only if the server runs without --token.

3. Run:

   python3 "${CLAUDE_PLUGIN_ROOT}/scripts/collect.py" init --name NAME --url URL --token TOKEN

4. Do the first sync and report the result. It reads every local session log once, so allow a minute on an old machine:

   python3 "${CLAUDE_PLUGIN_ROOT}/scripts/collect.py"

5. Tell them: from now on every session end syncs automatically, and the dashboard is at the URL they gave. Mention once that only counts are sent — session ids, token totals, timestamps, model names, project paths — never prompt or file content.

If step 4 fails with a connection error, the dashboard URL is wrong or the server is down; with bad token, the secret does not match the server's --token.