* fix(cli): wait for the old daemon to finish ongoing requests during restart The daemon shuts down gracefully: it stops listening right after /stop, but keeps serving ongoing requests before disposing of the wallet and releasing wallet.pid. The restart flows only waited for the health check to go down, then spawned a replacement daemon that failed to claim the routstrd wallet lock and exited with code 1, surfacing a confusing 'Cannot claim the routstrd wallet lock ... PID X is still running' error. Add waitForDaemonToExit() and use it in restart, mode, the post-update restart, and stop: - Phase 1: wait (10s) for the health check to stop responding. - Phase 2: while the old process still holds the wallet lock, show 'Finishing all ongoing requests...' (heartbeat every 10s) and wait up to 10 minutes for it to exit. A stale lock (dead PID) is not waited on; after the timeout the error names the holding PID and how to force it. stop now also waits for full exit instead of returning as soon as /stop is acknowledged, and no longer auto-starts a daemon when none is running. * fix(cli): offer 'kill -9 <PID>' to force stop a draining daemon The drain progress messages and the drain-timeout error now suggest 'kill -9 <PID>' so the user can force the old daemon out instead of waiting for stuck requests. A plain SIGTERM would not work: the daemon's signal handler re-runs the same graceful shutdown (server.close()), which keeps waiting for ongoing requests, so only SIGKILL can interrupt a stuck drain. The wait loop already treats the resulting dead-PID lock as released (same liveness semantics as claimPidFile), so the restart proceeds cleanly. The heartbeat interval is now injectable (drainHeartbeatMs, default 10s) like the other timing knobs, which also lets the tests cover the heartbeat message. --------- Co-authored-by: redshift <213178690+1ftredsh@users.noreply.github.com>
routstrd
Routstr daemon - A CLI tool for managing routstr processes, similar to cocod (a Cashu wallet daemon).
Overview
routstrd is a Bun-based CLI tool that provides a background daemon for the Routstr protocol. It integrates with cocod for wallet management and uses the Routstr SDK to handle provider routing and model discovery.
Routstr for Teams
For team-based routing, see routstrd-auth.
Features
- Daemon Mode: Run routstrd as a background HTTP server
- Wallet Integration: Works with cocod for Cashu token management
- Provider Routing: Automatically discovers and routes requests to available providers
- Config Management: Stores configuration in
~/.routstrd/
Requirements
- Bun runtime
curl -fsSL https://bun.com/install | bash
Installation
Step 1: Install
Global with bun:
bun i -g routstrd
OR - From source:
git clone https://github.com/routstr/routstrd.git
cd routstrd
bun install
bun link
Step 2: Setup & Fund
routstrd onboard
routstrd receive <cashu> # receive a Cashu token
routstrd receive 2100 # to top up 2100 sats with lightning
Step 3: Integrate with Claude Code
routstrd clients add --claude-code # or --pi-agent / --opencode
Use Routstrd Skill
Tip: You can also install the routstrd skill so the agent can manage routstrd for you.
More Commands
Start Daemon
Start the background daemon:
routstrd start
With custom port:
routstrd start --port 9000
The daemon binds to 127.0.0.1 by default. To expose it on another interface:
routstrd start --host 0.0.0.0
Only expose the daemon behind appropriate network controls.
With specific provider:
routstrd start --provider https://your-provider.com
CLI Commands
Check daemon status:
routstrd status
Get wallet balance:
routstrd balance
Test connection:
routstrd ping
Refresh models and client integrations on demand:
routstrd clients --manual-refresh # same as `routstrd refresh`
Turn the daemon's scheduled refresh on or off (no restart needed):
routstrd clients --disable-automatic-refresh
routstrd clients --enable-automatic-refresh
Stop the daemon:
routstrd stop
NPC (Lightning Address)
The in-process wallet registers the NPC (npubx.cash) plugin, which gives the daemon a persistent Lightning address backed by the wallet's Cashu mints. Payments to the address are imported into the wallet automatically (websocket push, plus manual sync on demand).
# Show your NPC Lightning address (username@npubx.cash, or npub fallback)
routstrd wallet npc address
# Claim a username (quote first, then confirm to pay the claim fee from the wallet)
routstrd wallet npc username myname
routstrd wallet npc username myname --confirm
# Manually sync paid NPC quotes into the wallet
routstrd wallet npc sync
Equivalent daemon endpoints: GET /wallet/npc/address,
POST /wallet/npc/username, POST /wallet/npc/sync.
Daemon API
The daemon exposes an HTTP server (default port 8008) with the following endpoints:
Health Check
GET /health
Automatic Refresh Settings
POST /settings/auto-refresh
Request body:
{ "enabled": false }
Enables or disables the scheduled refresh job. Persisted to the daemon's
config.json as autoRefresh.enabled and picked up on the next tick, so no
daemon restart is required.
Route Request
POST /
Request body:
{
"model": "model-id",
"messages": [...],
"stream": false
}
Response:
{
"choices": [...],
"usage": {...}
}
Wallet storage
The in-process Cashu wallet stores its mnemonic and proof database in
~/.routstrd/wallet/. On first startup, an existing wallet in ~/.cocod/ is
migrated automatically after routstrd verifies that the legacy cocod daemon is
not running. Back up your mnemonic before upgrading.
Set ROUTSTRD_WALLET_DIR to override the canonical wallet directory. The
COCOD_DIR, COCOD_SOCKET, and COCOD_PID variables are retained only for
locating and excluding a legacy external cocod process.
If both ~/.routstrd/wallet and ~/.cocod contain different wallets, startup
refuses to migrate rather than picking a mnemonic for you. Run
routstrd wallet doctor to compare the two wallets (mnemonic fingerprints,
timestamps, and balances) and see which one to keep.
Configuration
Configuration is stored in ~/.routstrd/config.json:
{
"port": 8008,
"host": "127.0.0.1",
"provider": null,
"cocodPath": null,
"autoRefresh": { "enabled": true }
}
autoRefresh.enabled (default true) controls the daemon's scheduled refresh
job, which re-fetches Nostr events, routstr21 models, and client integrations
every 21 minutes. Set it to false (or run
routstrd clients --disable-automatic-refresh) to turn the schedule off and
refresh manually with routstrd clients --manual-refresh. autoRefresh.intervalMs
overrides the 21-minute interval.
Environment Variables
ROUTSTRD_DIR- Config directory (default:~/.routstrd)ROUTSTRD_SOCKET- Socket path (default:~/.routstrd/routstrd.sock)ROUTSTRD_PID- PID file path (default:~/.routstrd/routstrd.pid)
Development
Install dependencies:
bun install
Run CLI:
bun run start
Run daemon:
bun run start
Typecheck:
bun run lint
Manual chat-completions smoke test
With a funded daemon running, create or reuse a client API key and pass one or more current model IDs to the smoke script:
routstrd clients add --name smoke-test
ROUTSTRD_API_KEY=<api-key> scripts/smoke/chat-completions.sh <model> [model ...]
Set ROUTSTRD_BASE_URL to test a daemon at a different address. The script
makes live provider requests that may spend wallet funds, so it is intentionally
not part of bun test.
Project Structure
routstrd/
├── src/
│ ├── index.ts # Entry point with shebang
│ ├── cli.ts # Commander CLI commands
│ ├── cli-shared.ts # IPC utilities
│ ├── daemon.ts # HTTP server daemon
│ └── utils/
│ └── config.ts # Path configuration
├── package.json
└── tsconfig.json
License
MIT