Top 200 ReposEXPLORER
Browse agents, categories and pages
Back to results
Planned, not doneThe source describes an intention, not a finished result.

Turn Langflow flows into MCP tools that Cursor or Claude can call

Langflow serves every project as an MCP server, so a flow you built visually, such as document Q&A, becomes a tool that Cursor, Claude Desktop or another MCP client can call by name.

Try the promptRead the original (opens in a new tab)
Upvote 0. Sign in to upvoteSign in to upvote · One vote per signed-in person. Site votes are separate from GitHub stars.

The problem

How it works

In Langflow each project gets its own MCP server, and flows with a Chat Output component can be exposed as its tools. From the project’s MCP Server tab you choose which flows to serve, rename each tool and write a description, then copy a ready-made JSON snippet into your client’s MCP config or use the auto-install option. The documentation’s example exposes a Document Q&A flow about one person’s CV as a tool; when the user asks Cursor about that person, the agent picks the tool, and when they ask about someone else it explains that the tool does not cover them. Servers use streamable HTTP with SSE as a fallback, and authenticated servers get an API key per project.

Results and limitations

This is Langflow’s documentation with an illustrative example, not a reported project, so it is marked planned.

Before you try it

Langflow stresses tool names and descriptions: by default a tool is named after the flow’s ID, which tells an agent nothing, and vague descriptions make it pick the wrong flow. Turn on authentication before exposing a server beyond your machine, since anyone who reaches it can run your flows and spend your model credits. Tool calls time out after 180 seconds unless you raise the limit. Results depend on your setup and were not independently reproduced by Top 200 Repos.

Tools and repositories

Tools
Langflow · uv (mcp-proxy) · Cursor, Claude Desktop or another MCP client
Prerequisites
A Langflow server with at least one flow that ends in a Chat Output component.

Try it yourself: Write a tool name and description for a flow

Fill the fields to adapt the prompt. Empty fields stay as visible [placeholders].

Preview
I am exposing a Langflow flow to coding agents as an MCP tool. The flow [What the flow does], using [Data it uses]. It should be used when [When to use it] and never when [When not to use it]. Suggest a snake_case tool name and a description of at most three sentences that tells an agent exactly what the tool covers, what input it expects and what it returns.

Ask the agent a question just outside the tool’s scope to check that it does not call it.

Related videos

About langflow-ai/langflow

About langflow-ai/langflow

Independent creators. Listing a video implies no endorsement, sponsorship or partnership.

Sources

  1. Use Langflow as an MCP server (opens in a new tab)

    Project documentation · docs.langflow.org · Undated · Source checked 12 Nov 2026, 18:05 UTC

    What this source supports
    • Langflow adds each project to its MCP server and makes the project’s flows available as MCP tools; flows need a Chat Output component
    • Flows can be enabled or disabled, renamed and described from the MCP Server tab
    • Unclear tool names and descriptions can cause agents to select tools incorrectly; the default tool name is the flow ID
    • Example of a Document Q&A flow used from Cursor, with the agent declining when the tool does not apply
    • Streamable HTTP with SSE fallback, API key authentication, JSON and auto-install client setup, and a 180-second default tool timeout
Report a problem

Planned, not done

Generate images on your own ComfyUI from Claude Code or Cursor with Comfy MCP

Comfy MCP, from the ComfyUI team, lets a coding agent or Claude Desktop run workflows and text-to-image jobs on the ComfyUI installed on your machine, check what nodes and models you have, and collect the outputs.

github.comComfy-Org/ComfyUI
Reported by source

Build a self-hosted AI agent workflow with human approval

Use n8n's visual canvas to wire an LLM agent to your apps, add tools and memory, and pause for human approval before it acts.

github.comn8n-io/n8n
Reported by source

Build an on-call investigation agent from your own alert history

Goutham Veeramachaneni connected Claude Code to Grafana's gcx CLI, his code and runbooks, then trained its skill on two weeks of past pages for his services.

gouthamve.dev · 21 Apr 2026anthropics/claude-code
Planned, not done

Hand CRM updates and follow-up drafts to Grok Bot agents with approval steps

Grok Bot launched in beta in August 2026: always-on agents with their own cloud computer that sign in to your apps, learn routines by watching you, and come back when something needs your approval.

x.ai · 11 Aug 2026
Planned, not done

Try a desktop-operating agent safely in Anthropic's computer use demo container

Anthropic's computer-use-demo runs Claude inside a Docker desktop that you watch in the browser, with a chat panel, so you can test screen, mouse and keyboard automation away from your own machine.

github.com
Reported by source

Apply the context engineering lessons from Manus to a long-running agent loop

The Manus team describes the context rules it settled on after rebuilding its agent framework four times: keep the prompt prefix cacheable, mask tools instead of removing them, and use files as memory.

manus.im · 18 Jul 2025