Workflow overview
Why this workflow matters
Helpful for business development and pipeline building.
📺 Full walkthrough video: https://youtu.be/QHvfykuSbF8 ⚠️ Self-hosted only — This template uses a community node (Firecrawl) and cannot run on AlekSystem Cloud. Who it's for This workflow is for community managers, SEO professionals, or team leads who want to deliver a curated, AI-written news digest to a Discord channel automatically — without manually browsing news sources every week. How it works A schedule trigger fires every Monday at 9am and fetches a list of watch topics from a NocoDB data table. For each topic, the workflow queries Google News via SerpAPI and extracts the top 3 article URLs. Each article is scraped using Firecrawl, and the raw markdown is cleaned before being passed to Claude Sonnet. Claude writes a structured French-language digest entry per topic, with media name, date, URL, and a 2–3 sentence summary. All topic digests are aggregated and split into Discord-safe chunks (under 2000 characters). An intro message is sent to Discord first, then each chunk is sent in a loop, followed by a closing conclusion message. How to set up [ ] Configure the Schedule Trigger with your desired run frequency [ ] Connect NocoDB credentials and point the table node to your watch topics table [ ] Add your SerpAPI API key to the Search Google News node [ ] Add your Firecrawl API key to the Scrape Article Content node [ ] Add your Anthropic credentials to the Claude Sonnet sub-node [ ] Connect your Discord bot and set the correct server and channel IDs Requirements NocoDB (or compatible data table) account SerpAPI account Firecrawl account Anthropic API key Discord bot with message permissions How to customize Swap NocoDB for Airtable or Google Sheets to manage your watch topics list Edit the AI prompt in the "Draft Watch Summary with AI" node to change the language, tone, or output structure of the digest Adjust the DISCORD_CHAR_LIMIT constant in the "Split Text for Discord" code node to handle different message size constraints
Best fit
Categories
Services
Use cases
Need another direction?