start / stop / restart / status

snippbot start

Start the Snippbot daemon server.

snippbot start                    # Start with defaults
snippbot start --port 9000        # Custom port
snippbot start --host 0.0.0.0     # Bind to all interfaces
snippbot start --dev              # Dev mode with auto-reload

Option	Default	Description
--port	18781 (from config / fallback)	API port
--host	"" — dual-stack bind (IPv4 + IPv6)	Host to bind. Client commands resolve "" to 127.0.0.1 when dialing.
--dev	off	Auto-reload on code changes

Once running, the daemon serves:

• API: http://{host}:{port}/api
• UI: http://{host}:{port}/
• Health: http://{host}:{port}/health

---

snippbot stop

Stop the running daemon.

snippbot stop                     # Graceful stop (SIGTERM, 15s timeout)
snippbot stop --force             # Force kill (SIGKILL)
snippbot stop --port 9000         # Stop daemon on custom port
snippbot stop --strict            # Exit 1 if no daemon was running

Option	Default	Description
--port	from config (default 18781)	Port the daemon is running on
--host	from config (default "" → 127.0.0.1)	Host to target. Symmetrical with start --host / status --host.
--force	off	Force kill instead of graceful shutdown
--strict	off	Exit non-zero if no daemon was running (script-friendly)

The graceful stop sends SIGTERM and waits up to 15 seconds for in-flight requests to complete. Use --force if the daemon is unresponsive.

By default snippbot stop exits 0 even when there was nothing to stop (so repeated invocations are idempotent). Pass --strict in scripts that need to distinguish “stopped one” from “none to stop”.

---

snippbot restart

Stop the running daemon, then start it again — one command instead of snippbot stop followed by snippbot start.

snippbot restart                  # Graceful stop, then start
snippbot restart --force          # Force kill, then start
snippbot restart --port 9000      # Restart the daemon on a custom port
snippbot restart --dev            # Restart in dev mode with auto-reload

Option	Default	Description
--port	from config (default 18781)	API port (used for both the stop and the start)
--host	from config	Host to target and bind
--dev	off	Start in dev mode with auto-reload
--force	off	Force kill the running daemon instead of a graceful stop

If no daemon is running, restart simply starts one. After stopping, it waits up to ~10 seconds for the port to free before starting; if the port is still held it exits 1 (retry with --force).

When the OS service is installed (snippbot service install) and no --dev / --port / --host override is given, restart goes through the service supervisor (same as snippbot service restart) so it doesn’t race the service’s own auto-restart. Otherwise the daemon runs in the foreground of your terminal, exactly like snippbot start.

---

snippbot status

Check whether the daemon is running.

snippbot status                   # Table output
snippbot status --json            # JSON output

Option	Default	Description
--json	off	JSON output
--port	18781	API port to check
--host	"" (dual-stack)	Host to check

Example output:

┌──────────┬──────────────────────────────┐
│ Property │ Value                        │
├──────────┼──────────────────────────────┤
│ Status   │ running                      │
│ Uptime   │ 2h 15m                       │
│ Host     │ 127.0.0.1                    │
│ Port     │ 18781                        │
│ API      │ http://127.0.0.1:18781/api   │
│ UI       │ http://127.0.0.1:18781       │
└──────────┴──────────────────────────────┘

JSON output:

{
  "running": true,
  "status": "ok",
  "host": "127.0.0.1",
  "port": 18781,
  "uptime_seconds": 8100
}

---

Health endpoints

The daemon exposes three health endpoints (no auth required):

Endpoint	Purpose
GET /health	Basic liveness
GET /health/ready	Readiness (databases loaded)
GET /health/live	Kubernetes liveness probe

curl http://localhost:18781/health
# {"status": "ok", "version": "0.5.0", "uptime_seconds": 3600}