Ask Telegram for a Release Digest: n8n, LangBot and GPT-6 Astra with Two Live Sources

Description and use

An automated n8n and LangBot workflow fetches GitHub Atom feeds, deduplicates and limits entries, asks GPT‑6 Astra for a source‑backed digest, and posts it to Telegram.

Problem

Manually checking multiple GitHub release feeds is time‑consuming and error‑prone.

Outcome

The pipeline automatically delivers a concise, source‑backed release digest to Telegram, eliminating manual copy‑paste and ensuring up‑to‑date information.

Workflow

• 1. Spin up n8n and LangBot using Docker Compose.
• 2. Create a webhook in n8n to receive a chat message from Telegram.
• 3. Add two RSS Feed Read nodes to fetch the Atom feeds of n8n and LangBot releases.
• 4. Run a Code node to deduplicate URLs, keep at most two entries per project, and truncate excerpts to 1,800 characters.
• 5. Call GPT‑6 Astra via an HTTP Request node, passing the prepared data and a prompt that asks for a source‑backed digest and distinguishes beta from stable releases.
• 6. Process the model response in a Code node to format a reply for LangBot.
• 7. Respond to the original webhook with the formatted JSON so LangBot can post the digest in Telegram.

Evidence

The workflow fetches the public GitHub Release Atom feeds for n8n and LangBot, keeps two entries from each project, deduplicates the links, and asks GPT-6 Astra for a source-backed digest.

Source

Create a Header Auth credential in n8n: Name is Authorization, Value is Bearer YOUR_API_KEY.

Source