Nostr2tg

Bridge Nostr long-form notes (kind 30023) into a Telegram channel. Fetch authors from a static list or a follow list (kind 3), format posts, and publish. Supports dry-run for safe testing.

Configuration

Environment variables:

• N2TG_TG_BOT_TOKEN (required): Telegram bot token
• N2TG_TG_CHAT_ID (required): Target chat/channel ID (e.g. -100xxxxxxxxxx)
• N2TG_MODE: authors or followlist (default: authors)
• N2TG_AUTHORS: comma-separated author pubkeys (when authors mode)
• N2TG_FOLLOWLIST_PUBKEY: pubkey to read follow list from (when followlist mode)
• N2TG_NOSTR_RELAYS: comma-separated relay URLs
• N2TG_TG_PREFIX: prefix line for Telegram posts (default: New Nostr article:)
• N2TG_LINK_NADDR_BASE: base URL for naddr links (default: https://njump.me/)
• N2TG_LINK_NIP05_BASE: base URL for NIP-05 links
• N2TG_SYNC_INTERVAL_MS: sync interval in ms (default: 3600000)
• N2TG_TG_POLL_TIMEOUT_MS: Telegram long-poll timeout in ms (default: 10000)
• N2TG_DRY_RUN: set to 1 to simulate sends (no Telegram posts)
• N2TG_SYNC_ALL_ON_EMPTY: when no pinned message exists, set 1 to backfill from baseline 0 (epoch). If unset/0, publishing halts until a pinned message is set.

Local development

mix deps.get
mix compile
MIX_ENV=dev mix dialyzer

Run with env vars set. For dry-run:

export N2TG_TG_BOT_TOKEN=xxx
export N2TG_TG_CHAT_ID=-100xxxxxxxxxx
export N2TG_DRY_RUN=1
iex -S mix

Docker

Build and run locally:

docker build -t nostr2tg:local .
docker run --rm \
  -e N2TG_TG_BOT_TOKEN=xxx \
  -e N2TG_TG_CHAT_ID=-100xxxxxxxxxx \
  -e N2TG_DRY_RUN=1 \
  nostr2tg:local

Deploy to Render.com

This repo includes render.yaml and a multi-stage Dockerfile.

1. Push the repo to GitHub.
2. In Render, create a new service → Worker → “Use existing render.yaml”.
3. Configure environment variables:
   • N2TG_TG_BOT_TOKEN (secret)
   • N2TG_TG_CHAT_ID (secret)

• Optional: N2TG_MODE, N2TG_AUTHORS, N2TG_FOLLOWLIST_PUBKEY, N2TG_NOSTR_RELAYS, N2TG_TG_PREFIX, N2TG_LINK_NADDR_BASE, N2TG_LINK_NIP05_BASE, N2TG_SYNC_INTERVAL_MS, N2TG_DRY_RUN.

4. Deploy. The worker process runs /app/start.sh which launches the Erlang release.

Notes

• Dry-run logs the full flow and skips Telegram network calls.
• Baseline initialization and pinning:
  • On startup, baseline is derived from the Telegram channel's pinned message.
    • If the pinned text ends with a line like Published: YYYY-MM-DD HH:MM:SS (UTC), that timestamp is used.
    • Otherwise, the pinned message's own date is used as baseline.
  • If there is no pinned message:
    • With N2TG_SYNC_ALL_ON_EMPTY=1, baseline is 0 (backfill everything).
    • Otherwise, publishing halts until a message is pinned.
  • After each batch, the last sent message is pinned to keep the baseline current.