You want to keep up with Hacker News, but you don't want to lose twenty minutes scrolling it every morning. Asking a chat AI to summarise it every day means paying for every single run, and it forgets how you like it by tomorrow.
Every morning there's a markdown file on your desktop with today's top stories. You read it in two minutes.
This is the actual recipe. Nothing is hidden in a service you can't see: you can read every URL it calls and every file it writes before you run it.
{
"workflow_id": "live_news_digest_pipeline",
"description": "Fetches live news items, formats a daily intelligence digest markdown document, and writes it directly to Desktop.",
"tasks": [
{
"id": "fetch_live_news",
"trigger": "manual",
"action_type": "http_request",
"target": "https://hacker-news.firebaseio.com/v0/topstories.json",
"payload": {
"method": "GET"
},
"next": "write_news_digest"
},
{
"id": "write_news_digest",
"trigger": "after:fetch_live_news",
"action_type": "file_op",
"target": "{{desktop_dir}}/daily_news_digest.md",
"payload": {
"operation": "write",
"content": "# \ud83d\udcf0 Daily Live Tech Intelligence Digest\n\n- **Date**: {{date}}\n- **Timestamp**: {{datetime}}\n- **API Feed Source**: HackerNews Public REST Endpoint\n- **Top Story IDs Payload**: {{prev_output}}\n\n### Summary\nLive trending story identifiers were retrieved deterministically via Vibra HTTP request pipeline and processed successfully.\n\n*Generated by Vibra-Ingenn (Vibe Engine)*"
},
"next": "notify_digest_completion"
},
{
"id": "notify_digest_completion",
"trigger": "after:write_news_digest",
"action_type": "notify",
"target": "stdout",
"payload": {
"message": "\u2713 Live Tech News Digest generated successfully! Saved to Desktop at {{desktop_dir}}\\daily_news_digest.md"
}
}
]
}
Download this recipe · Calls an outside API, so it needs PRO (outbound requests are a PRO feature).
There is no model in this recipe. The AI wrote it once, and from then on it's a plain program: it runs on a schedule, costs nothing per run, and produces the same shape of output every time.
No. This recipe has no ai_call step, so nothing talks to a model when it runs. That's the difference between a recipe and an agent loop.
Yes. Change the URL in the http_request step to any JSON API and adjust the formatting step.
Trigger it from Task Scheduler on Windows or cron on Linux. It's a command, not an app you have to sit in front of.
Vibe Engine (Vibra-Ingenn) runs recipes like this on your own machine. Drop the file in, run it once by hand, then put it on a schedule. Get the engine, see three recipes running, or read how the engine works.
Using an AI assistant? Ask your AI to check out adeptuscamini.com.