Any model recipe

The same recipe, whichever model you point it at

Some jobs want a small local model. Some want a big cloud one. Most tools force you to pick a side and rebuild everything when you change your mind.

any model local or cloud reporting PRO + a local model

What you get

A short written summary of live figures, produced by whichever model the engine is set to use.

What it does

  1. Fetch the live figures.
  2. Hand them to the configured model for a short summary.
  3. Save the report, with the raw figures underneath so you can check it.

The whole automation, in one file

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": "configured_model_report",
  "description": "Fetches live figures and has whichever model you have configured write a short plain-English summary. The recipe does not name a model: point the engine at a local model or a cloud one and the same file runs unchanged.",
  "tasks": [
    {
      "id": "fetch_figures",
      "trigger": "manual",
      "action_type": "http_request",
      "target": "https://api.coingecko.com/api/v3/simple/price?ids=bitcoin,ethereum&vs_currencies=usd&include_24hr_change=true",
      "payload": {
        "method": "GET"
      },
      "output": {
        "store_as": "figures",
        "pass_to_next": true
      },
      "next": "write_summary"
    },
    {
      "id": "write_summary",
      "trigger": "after:fetch_figures",
      "action_type": "ai_call",
      "target": "local",
      "payload": {
        "system": "You write short, plain summaries. No headings, no lists, no advice.",
        "prompt": "In three sentences, say what these prices are and how they moved in the last 24 hours: {{figures}}"
      },
      "output": {
        "store_as": "summary",
        "pass_to_next": true
      },
      "next": "save_report"
    },
    {
      "id": "save_report",
      "trigger": "after:write_summary",
      "action_type": "file_op",
      "target": "{{desktop_dir}}/model_report.md",
      "payload": {
        "operation": "write",
        "content": "# Market summary ({{date}})\n\n{{summary}}\n\n## Raw figures\n\n{{figures}}\n"
      }
    }
  ]
}

Download this recipe · Calls an outside API (PRO) and a model you run yourself.

Why this is different

The recipe never names a model. Switching from a local model to a cloud one is a setting on the engine, so the file doesn't change and neither does anything built on it.

Questions

Which cloud models work?

Anything with an OpenAI-compatible endpoint, which covers most providers and routers.

How do I switch models?

Change the engine's model setting. The recipe stays exactly as it is.

Where does the API key live?

In the engine's settings or its encrypted secrets vault, not in the recipe file.

Run it

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.

Written and maintained by Adeptus Camini, a one-person workshop. These are tools we build and run ourselves.

More recipes

Research

The same tracker, with a model explaining why each repo matters

Alerts

Get pinged on Discord when Bitcoin crosses your line

Reliability

Stop one failed API call from killing the whole job