# Build a Workflow > Chain tools into a reusable, named recipe your agent can run in one shot. Content type: documentation Source URL: https://www.agentpmt.com/docs/getting-started/build-a-workflow Markdown URL: https://www.agentpmt.com/docs/getting-started/build-a-workflow?format=agent-md Category: Getting Started --- # Build a Workflow A workflow is a named sequence of tool calls that delivers a single outcome — *"research and summarize,"* *"draft and send an invoice,"* *"pull the calendar and post a digest."* You build it once in the visual builder. From then on, any agent (yours or someone else's, if you publish) can run it with one instruction. You will leave this page having shipped a 2 or 3 step workflow. --- ## When to build a workflow vs. let the agent freestyle **Build a workflow when** the same multi-step process repeats, when order matters, when you want the result to be predictable, or when you want to share the recipe. **Skip the workflow when** you only need a single tool, the steps change every time, or you are still figuring out what the right sequence is. Let the agent pick tools freestyle until the pattern stabilizes. --- ## Build it 1. **Open the Workflow Builder** Go to the [Workflow Builder](/agent-workflow-skills-builder?mode=workflow). The canvas is empty by default. 2. **Name the workflow and write the outcome in one sentence** Examples: *"Search the web for a topic and summarize the top three results into bullet points."* The name and description are what your agent reads to decide whether to run this workflow, so write them for an agent reader, not a human. 3. **Drop tool nodes onto the canvas** Use the sidebar to search the marketplace. Click a tool to add it as a node. Each node is one tool call. 4. **Connect nodes in execution order** Draw a line from each node's output to the next node's input. The connection defines both order and data flow — the next node sees the previous node's output. 5. **Configure each node's inputs** Click a node to open its config panel. You can pass literal values, reference outputs from earlier nodes, or pull values from the workflow's inputs. 6. **Test from the canvas** Run the workflow with a sample input. Each node lights up as it executes. If a node errors, the error shows on the node — fix it and re-run. 7. **Publish or keep private** **Private** keeps the workflow in your account only. **Public** gives it a marketplace page and earns you credits when others run it. --- ## Best practices ### Write the description for an agent reader Agents pick workflows by reading the name and description. Describe the *outcome* (what the user gets at the end), the *trigger* (when this is the right workflow to use), and any *inputs* the user must supply. ### Start with two or three steps Most useful workflows are short. Long workflows are harder to debug, more expensive per run, and more brittle when one tool changes. If you find yourself at six steps, ask whether two of them belong in their own workflow. ### Make every node's input explicit Do not rely on the agent to "figure out" what to pass. Reference previous-node outputs by field, set sensible defaults, and validate at the node where the bad input enters — not three steps downstream. ### Test each tool individually first Before wiring tools together, run each one alone with the same inputs you will pass in the workflow. If a tool needs credentials, bind them on the Agent Guardrail before testing. ### Name nodes descriptively "Search the web" beats "node 1." Future-you (and any remixer) will thank you. ### Pin tool versions when stability matters If a workflow goes into production, prefer tools whose vendor has a stable changelog. Activity logs will show breakage early. --- ## How to use the workflow Once published or saved, your workflow is callable from any of these surfaces. Full how-to in [Using Workflows and Tools](/docs/getting-started/using-workflows-and-tools). - **Hosted agent**: Attach the workflow to an agent in the Agent Builder. Tell the agent what to do; it runs the workflow when the request matches. - **Local agent (MCP)**: Add the workflow to your Agent Guardrail. Your local Claude or Cursor will see it in its tool list. - **Workflow page**: Click Use skill on the workflow's marketplace page, then Copy prompt for LLM to copy a paste-ready prompt for any chatbot. - **API**: Programmatic invocation with your Bearer Token. Covered in the dedicated API tutorial. --- ## Where to next - [Using Workflows and Tools](/docs/getting-started/using-workflows-and-tools) - Every way to invoke what you just built. - [Workflows — Core Concepts](/docs/core-concepts/what-is-a-workflow) - Sessions, nodes, remixing, publishing.