Skip to main content

Quickstart

Need an agent in five minutes? Weaver is the no-code AI workflow builder inside SmythOS.
Describe a goal in plain language → Weaver assembles, connects, and generates the agent for you.

TL;DR
In the next three minutes you’ll create, debug, and deploy a working agent without touching code, APIs, or the visual canvas.

Prerequisites​

  • A SmythOS account (sign up if you’re new).
  • Zero coding skills. Weaver handles the heavy lifting.
  • Think of a goal... e.g. “Send daily summaries from Airtable to Slack.”

That’s it. No coding. No hassle.

Step 1 - Describe What You Want​

Open Weaver and paste the prompt below:

Create an agent that writes SEO-optimized blog posts and publishes them to WordPress. The agent should have the following skills: - Write article from a topic - Extract research keywords - Post article to WordPress using a given URL

Weaver will instantly:

  • Parse the prompt
  • Ask questions to help tailor the workflow
  • Provide a plan after thinking
  • Select and connect relevant components
  • Build the first draft visually

Step 2 - Let Weaver Think and Propose​

After you describe your goal, Weaver enters a "Thinking..." phase where it:

  • Interprets your intent

  • Suggests a workflow plan based on your request

  • Asks follow-up questions to clarify details like:

    • What kind of blogs are you targeting?
    • Do you want keyword extraction to be optional or required?
    • Which CMS platform are you using—WordPress.com or self-hosted?

Then it drafts a complete workflow such as:

  • Generate blog structure
  • Write full content
  • Research topics
  • Publish (optional)

Step 3 - Review the Draft​

You’ll see a canvas with three cards:

  • GenAI Writer
  • Keyword Extractor
  • WordPress Poster

Hover to inspect inputs/outputs, click to edit, or drag to adjust wiring.

Circles legend -> ● = required input, ○ = optional

Step 4 - Test It Live​

Ask the agent:

Write a post about “Top AI trends in 2025.”

Watch real-time outputs:

  • Article text
  • Extracted keywords
  • WordPress publish confirmation

Weaver’s LLM orchestration framework gives inline feedback, no silent errors.

Step 5 - Enable Debug Mode​

If something goes wrong:

  1. Click any block with a red border to open the error details
  2. Review the input/output snapshot and Weaver’s explanation
  3. Hit "Fix with AI" — one click usually solves it
  4. Cmd/Ctrl + click to select and fix multiple blocks at once

Step 6 - Deploy Your Agent​

When everything passes:

  1. Choose Deploy and go with the default Agent Cloud option which hosts the agent on our infrastructure or download the SRE for running locally.
  2. Tag your version in case you want to restore a previously deployed instance.

Weaver allows you:

  • Share the agent to collaborate with your friends and colleagues by inviting them
  • To debug using activity logs
You did it!
You used prompt-based agent creation to build, test, and deploy without ever opening the drag-and-drop canvas. That’s true no-code agent building.