BitMacro Relay Manager

Running a Nostr relay in production today usually means bash scripts, raw CLI, and almost no operational visibility.

# This is what managing a relay looks like today
strfry scan --filter '{"authors":["abc123..."]}' | jq .
nak req -k 0 wss://relay.example.com | grep spam
vim /etc/strfry/whitelist.txt && systemctl reload strfry

Relay Manager eliminates this. A visual panel, a central API, a stateless agent on your box, and documented flows for NIP-07 (browser extension) and NIP-46 (remote signer) where users need to sign without pasting keys into the web app.

The public product story and CTAs live on bitmacro.io/relay-manager. These pages are the technical companion: architecture, modules, protocol notes, and a consolidated API reference.

The problem space

Operators typically juggle:

• Whitelists edited by hand on the server
• Spam and moderation via CLI-only tools
• No single place for disk usage, uptime, or kind breakdown
• Paid relay flows reinvented per deployment

There is no maintained, relay-agnostic visual tool in the vein of abandoned strfry-web-ui. Relay Manager targets strfry first (broad deploy base) with a clean agent API so other backends can follow later.

Ecosystem

Five public-moving parts plus one private hub:

Roadmap: @bitmacro/connect will centralise /signer and NIP-46 helpers for panel, connect, and third-party apps; until then @bitmacro/relay-connect is the npm surface (see relay-connect).

Authentication: three layers you should not confuse

Panel login is not NIP-07. Connect flows are where NIP-07 and NIP-46 ship today — see NIP-07 and NIP-46.

Core Principle

The panel never speaks directly to the agent or to Supabase.
All orchestration goes through the Web Server (relay-api).

Browser
  │
  ▼
relay-panel (Next.js)          ← UI only; zero direct DB/agent access
  │  X-API-Key + X-Provider-User-Id
  ▼
relay-api [Web Server]         ← Supabase CRUD + agent proxy + /signer NIP-46
  │  Bearer token per relay
  ▼
relay-agent [Relay Server Agent]   ← strfry CLI abstraction; stateless
  │
  ▼
strfry / nostr-rs-relay        ← the actual relay process

What v0.2.0 Solves

Current Production State

BitMacro operates three relays on a single Relay Server Agent (port 7810):

Health monitoring: Uptime Kuma on https://relay-agent.bitmacro.io/health — HTTP keyword "status":"ok".

Licenses and open source strategy

BSL code can still be self-hosted; the hosted panel on BitMacro optimises for convenience.

Roadmap (condensed)

Done (v0.2.x)

• Multi-relay Relay Server Agent — one process, RELAY_INSTANCES, /:relayId/… API
• relay-api CRUD + proxy and /signer + nip46_sessions
• relay-panel 0.2 — create relay with agent_relay_id, full dashboard
• relay-connect-web — server-side proxy, NIP-46 QR flow, NIP-07 extension path

Next

• v1.0 — Lightning — invoice, webhook, auto-access after payment (LNbits / LND family)
• @bitmacro/connect npm package — shared /signer client
• Optional: nostr-rs-relay backend, metrics, richer alerts

Where to read next