AI Workflow Builder

You can build a complete workflow - nodes, connections, and configuration - by describing what you want in plain language. Building from scratch is handled by the AI Chat Assistant right in the workflow editor.

How to Use

1. Create a new workflow and choose Blank.
2. The editor opens with the AI Assistant panel already active: "Let's build your workflow" with the input placeholder "Describe your workflow…".
3. Describe what you want to automate. The assistant grounds itself in your workspace context (available AI models, MCP servers, knowledge bases, forms, and global variables), so it only proposes things that can actually run.
4. If it needs more information, it asks a focused clarifying question - answer and continue.
5. The assistant first shows a plan card summarizing the workflow it intends to create. Review it and click Build it.
6. The assistant then adds nodes and connections as action cards in the chat - accept them one by one or click Apply all to apply the whole batch. Nothing is applied without your confirmation.
7. Before declaring the workflow ready to run, the assistant validates requirements. Anything missing - an AI provider key, an MCP server, a global variable - appears as an inline setup card you can fill in without leaving the chat.

You can also open the assistant on any existing workflow (the Sparkles icon in the toolbar) and extend it the same way: "add a Slack alert when the check fails".

Workspace Context

The assistant automatically detects the resources available in your current workspace:

• AI Providers - which models the Agent nodes can use
• MCP Servers - available tools and their capabilities (e.g., Slack messaging, stock data)
• Knowledge Bases - document collections available for RAG queries
• Forms - available as form-trigger candidates
• Global Variables - workspace key-value store entries

This means the generated workflow only references models and tools that are actually configured. If your request needs a capability that is not configured yet (e.g., you ask for Slack notifications but no Slack server is set up), the assistant flags it and offers an inline setup form instead of silently generating a broken workflow.

Tips for Good Prompts

Be specific about the trigger

Tell the assistant how the workflow should start:

• "Create a webhook that receives order data..."
• "Run every day at 9am..."
• "Manual trigger that accepts a list of URLs..."

Describe the data flow

Explain what happens at each stage:

• "Fetch the stock prices, then analyze trends with AI, then send a summary to Slack"
• "Receive a support ticket, search the knowledge base for relevant docs, generate a response"

Mention the output format

If you need structured output, say so:

• "Return the result as JSON with fields: status, summary, and action_items"
• "Send a formatted report to the #analytics Slack channel"

Examples of effective prompts

• "Create a workflow that monitors stock prices for AAPL and TSLA every hour. If any stock moves more than 5% from the previous check, send a Telegram notification with the details."
• "Build a webhook handler that receives customer feedback, analyzes sentiment using AI, and stores the result in a Google Sheet. Use the knowledge base to find relevant product documentation for context."
• "Make a daily scheduled workflow that fetches open YouTrack issues, groups them by priority, generates a summary report with AI, and posts it to the #dev-updates Slack channel."

Limitations

• The assistant builds the workflow but does not execute it. Review the generated nodes and adjust parameters as needed before running.
• Very complex workflows with many branching paths may need manual refinement - build the core flow first, then iterate node by node in the same chat.

For the full reference on modes, action cards, slash commands, and usage limits, see AI Chat Assistant.