doctor

Doctor

Health check for this agent. Inspects config, identity, memory, SQLite index, QMD, crons, hooks, HTTP bridge, messaging plugins, and dreaming. When invoked with --fix, applies safe auto-repairs (create memory dir, sync index, delete stale BOOTSTRAP.md) before re-running checks.

This is a CORE feature — always available. See docs/doctor.md for the full list of checks and auto-fixes.

When to use

• Right after /agent:create or /agent:import — verify setup is clean
• When memory search returns nothing or feels wrong — check SQLite integrity
• When heartbeats don't seem to fire — verify crons and active hours
• When you enabled the HTTP bridge but the browser can't connect — probe it
• Anytime something feels off

How it works

1. Refresh the cron audit FIRST — CronList → writeback.sh audit — so the registry's .audit state is current
2. Call the MCP tool agent_doctor with action='check' (or action='fix' when the user passes --fix); its cron-registry check reads the .audit you just refreshed
3. The tool returns a structured report; pass it through as-is (the tool already formats a card)
4. Supplement with cron status rendered from the same CronList output — the MCP server cannot see Claude Code's cron state

Steps

Step 1 — Detect mode

• If the user's message contains --fix or /agent:doctor fix or says "arreglar" / "auto-fix" / "repara" → mode is fix
• Otherwise → mode is check

Step 2 — Refresh the cron audit (BEFORE the card)

The agent_doctor card reads the registry's persisted .audit offline. Refresh it first, or the card may render stale/contradictory cron state.

CronList is a deferred tool. Load it first:

ToolSearch(query='select:CronList')

Call CronList() and pipe its FULL output (verbatim, including a literal No scheduled jobs.) to the mechanical auditor:

printf '%s\n' "<full CronList output>" | bash "$CLAUDE_PLUGIN_ROOT/skills/crons/writeback.sh" audit

(Exit 4 = CronList format drift — remember it and report it as its own ⚠️ line in Step 4.) Keep the CronList output for Step 4.