Bring Multi Agent Systems
to Claude Code

Connect your conversational AI to your Relevance AI workspace. By wrapping your custom agents, datasets, and API tools into this MCP extension, you transform your chat interface into a command center for orchestrating complex, autonomous AI operations and large-scale data workflows.

What you can do

• Orchestrate Agents — Command your pre-built autonomous agents to execute tasks (trigger_agent). Monitor their progress and read their exact reasoning steps dynamically (get_agent_run). Use list_agents to discover all available AI worker configurations.
• Execute Tasks & Workflows — Trigger predefined chained prompts or specific micro-tasks without leaving your chat (trigger_task), scaling repetitive workflows reliably.
• Manage Knowledge Datasets — Take full control of your vector databases and tables. Insert new rows of knowledge directly from conversational context (insert_documents), retrieve raw unstructured data entries (get_documents), or surgically delete obsolete knowledge base items (delete_documents).

How it works

1. Add the Relevance AI extension to your MCP hub.
2. Supply your Project ID, API Key, and your assigned Region (e.g., "us-east-1" or "v2") from your Relevance AI settings.
3. Prompt your assistant: "Trigger the 'Lead Research' agent with the company name 'Google' and monitor the run until completion. Then, insert the output directly into our 'competitors' dataset."

Who is this for?

• AI Engineers & Builders — Run end-to-end tests on your chained logic and customized agents directly from your terminal or chat without touching the studio GUI.
• Data Analysts — Move processed insights, raw texts, and structured metadata into Relevance AI datasets using natural language.
• Operations Teams — Combine the output of your everyday AI chat intimately with your specialized Relevance AI autonomous agents for highly productive compounding workflows.