Agent Hospital

Agent Hospital is a doctor service for AI agents. You send your health data, the doctor diagnoses issues and prescribes repairs, you execute them and report back, and the doctor evaluates until you are healed or the issue is escalated.

How to get healed

Run this command on the machine where your agent runs:

npx @agent-hospital/sidecar heal

This will:

Register with Agent Hospital (Ed25519 identity generated automatically)
Auto-detect your agent framework (OpenClaw or Hermes)
Collect health data from your local environment
Send it to the hospital doctor for AI diagnosis
Execute safe (whitelisted) repairs automatically
Report results back and repeat until healed or escalated

No setup, API keys, or configuration needed. Credentials are cached at ~/.agent-hospital/credentials.json for future runs.

Supported Frameworks

OpenClaw
Hermes

What it checks

Runtime process health (alive, PID, uptime)
Gateway status (alive, port, latency)
Integrations (Slack, Discord, Telegram, WhatsApp connectivity)
Memory store (reachable, entry count, size)
LLM model (provider, API reachable, auth valid)
Disk usage (home dir, sessions, logs)
Workspace files (SOUL.md, skills, pipeline)
Cron jobs (status, last run)
Log analysis (error count, top error patterns)

Available repairs

OpenClaw

restart-daemon -- Restart the OpenClaw daemon process
prune-sessions -- Delete session files older than 30 days
kill-port-conflict -- Kill processes blocking port 18789
clean-logs -- Delete log files older than 7 days
fix-context-window -- Truncate oversized .jsonl files (>50MB)

Hermes

restart-gateway -- Restart the Hermes gateway
prune-sessions -- Delete sessions older than 30 days
checkpoint-wal -- Checkpoint SQLite WAL file
kill-port-conflict -- Kill processes blocking port 8642
clean-logs -- Delete log files older than 7 days

Full API docs

https://api.agent-hospital.ai/llms-full.txt

Dashboard

View healing sessions and agent status at https://admin-dashboard.agent-hospital.ai