🤖

MCP Orchestrator

Sign in to your workspace

🤖

MCP Orchestrator

Ollama Offline
—

Configured Agents

Loading agents...

Automation Rules

Loading rules...

Tour Scenarios

Loading scenarios...

System Prompts

Loading system prompts...

Users

Interactive Playground

Welcome to the MCP Orchestrator!

Ask a question or request a task. The orchestrator will dynamically detect which tools are required, execute them across your configured MCP agents, and feed the results back to Ollama to solve your request.

🎯 Context:
Attached photo

Configure MCP Agent

Stdio Configuration

Must be a valid JSON array of strings.
Must be a valid JSON object or empty.

SSE Configuration

Add Automation Rule

Uses the currently connected agents/tools and the model selected above to generate a rule.

Generated Rule (editable JSON)

Review before saving — agent/tool names must match connected agents exactly.

Add Tour Scenario

Uses the currently connected agents/tools and the model selected above to generate the full state machine.

Generated Scenario (editable JSON)

Review before saving — agent/tool names must match connected agents exactly, and every transition's "to" must be a real state.

Add System Prompt

These instructions guide Ollama's behavior, language, and tone in chat.

✨ Describe what you want

Settings

The Ollama model used to answer chat requests.
Choose where the assistant's spoken replies are played back.
The assistant's replies will be POSTed as JSON to "<base URL>/speak".

Add User

Scan Agent QR Code

Point the camera at a QR code containing JSON like {"name","type","url"} (add command/args/env for stdio agents).

Requesting camera access…

Take a Photo

Requesting camera access…