FLUJO

Building an AI agent in FLUJO means saving a model connection once, adding plain-language steps or dragging nodes onto a visual graph, and attaching tools from any MCP server, all inside one open-source workspace you host yourself. Connections cover OpenAI, Azure OpenAI, Anthropic, Google Gemini, xAI, OpenRouter, and local Ollama models, with ordered fallback connections for rate limits and outages plus a built-in connection test. The Connected Apps screen imports MCP servers from curated apps, the MCP Registry, a remote URL, a GitHub repository, or a local command, then lets you call their tools and read their resources before an agent touches them. Easy mode assembles an agent one recipe step at a time, Expert mode exposes Start, AI, connected-app, Static, and Subflow nodes with branching handoffs and loops, and both open the same saved agent. Talk runs agents in a live chat with a context-window meter, optional tool approvals, file and audio attachments, message editing, and a visual debugger that sets breakpoints and steps through nodes. Triggers start flows from schedules, webhooks, file changes, polled MCP tools, changed websites, or another flow finishing. Experimental Personas keep memories and goals between sessions, and Meetings seat several agents in one conversation. Any saved flow answers as a model on an OpenAI-compatible /v1 endpoint, so Cline, Cursor, or SDK clients can call it. Running on a dedicated VPS on RepoCloud with guaranteed CPU, RAM, and SSD, full root SSH access, and a browser serial console. MIT licensed.

FLUJO
FLUJO
FLUJO
FLUJO
FLUJO
FLUJO
FLUJO
FLUJO

Benefits

  • Build Agents Like Writing Recipes
  • Easy mode turns an agent into plain-language steps with a model and tools bound to each one, and the same saved agent opens as an Expert graph with branching, loops, and subflows.
  • Switch Models Without Rewiring Agents
  • Save OpenAI, Anthropic, Gemini, OpenRouter, or Ollama connections once, test each with a built-in round-trip check, and add ordered backup connections that take over during rate limits or outages.
  • Call Your Agents From Any Client
  • Saved flows answer as models on an OpenAI-compatible /v1 endpoint, so Cline, Cursor, and SDK clients can run them. A built-in /docs page documents the chat, flow, MCP, and automation endpoints.
  • Automate Without Keeping A Tab Open
  • Schedule, webhook, folder watcher, tool poller, website change, and flow-completion triggers run agents in the background, and the server keeps working after the browser tab closes, with run history per trigger.

Features

  • Visual Agent Builder
  • Easy recipe steps and an Expert graph of Start, AI, connected-app, Static, and Subflow nodes edit one saved agent, with branching handoffs, loops, and per-step model and tool scope.
  • MCP Connected Apps
  • Import MCP servers from curated apps, the MCP Registry, a remote URL, a GitHub repository, or a local command, then test tools and render interactive MCP Apps in chat.
  • Debugger And Tool Approvals
  • Breakpoints, step-through execution, and state inspection sit beside a context-window meter, live token usage, and optional approval prompts that pause a run before a tool executes.
  • Automation Triggers
  • Six trigger types start flows: schedules, authenticated webhooks, watched folders, polled MCP tools, website changes, and completion of another flow, each with its own run history.
  • Personas And Meetings
  • Experimental Personas keep memories, conversations, tasks, and goals across sessions, and Meetings seat several agents in a shared conversation with participants and facilitation chosen in a four-step wizard.
  • Docker Compose Deployment
  • docker compose up --build serves the app on port 4200 with named volumes for workspaces, database, and MCP servers; put an authenticating reverse proxy in front for network access.

Apps Similar to FLUJO