Inrō can build a complete scenario, triggers and actions included, from a plain-language description. You describe what you want the automation to do, and the AI handles the configuration. This is available through the API and the MCP server.
Why use it
Instead of selecting triggers and adding actions one by one, you describe your goal in plain text and get a ready-to-review scenario in seconds. It's the fastest way to go from idea to draft.
How it works
Create a scenario with a
promptdescribing what the automation should do.The scenario is created right away, and the AI generates its triggers and actions in the background.
Poll the scenario to track progress. The
generation_datafield reports a status for each phase,triggersandactions, aspending,generating,completed, orfailed.Once both phases are
completed, the scenario is ready to review and activate.
From your prompt, Inrō also detects whether the scenario should be a custom flow or a comment-to-DM flow, and builds it accordingly.
Using the API
POST /api/v1/scenarios{ "prompt": "Reply to comments on my post with a DM containing a link" }
The response returns a 201 immediately with the new scenario. Check progress with:
GET /api/v1/scenarios/:id
The generation_data field shows per-phase status. Once both phases complete, triggers and actions are populated in the response.
Using the MCP server
The create_scenario MCP tool accepts the same prompt argument, so you can build a scenario from ChatGPT or Claude in plain language. Use get_scenario to poll for completion.
What to know before you rely on it
Review before activating. Always read the generated scenario before turning it on. A scenario can't be activated while it's still generating.
promptandtemplate_idare mutually exclusive. Use one or the other, not both.The AI builds from a focused set of steps. Generation uses common triggers (incoming DMs, comments, story replies, and story mentions) and a core set of actions (send message, delay, ask a question, collect a property, button message, conversion link, like, and folder). For anything outside that set, add the step yourself in the builder afterward.
Generated flows start without variables. The AI leaves personalisation to you, so add
{{contact.name}}and other variables after reviewing.
🐾 Netsuke's Tips
Be specific in your prompt. "DM the link to anyone who comments LINK on my latest post, then follow up in a day if they don't click" produces a far better draft than "comment to DM."
Treat the result as a first draft. Generation gets the structure right; you refine the copy, links, and timing.
Poll
generation_datarather than assuming the scenario is ready the moment it's created. The triggers and actions fill in a few seconds later.
What's next?
To change a generated scenario later by editing its prompt, read Update a Scenario with a New Prompt (AI Regeneration). For the wider API and MCP surface, see Build and Update Automations Programmatically via API & MCP and The Inrō Private API: Getting Started.