Platform — MCP

Your AI tools. Redbird's agents and workflows. Connected.

Redbird MCP lets your AI tools — Claude, ChatGPT, and others — reach directly into Redbird's AI agents, automation workflows, and data outputs through the Model Context Protocol. Ask your AI to run a workflow, query a dataset, or trigger an analysis. Redbird handles the execution.

claude_desktop_config.json JSON · MCP
{
  "mcpServers": {
    "redbird": {
      "command": "npx",
      "args": ["-y", "@redbird/mcp"],
      "env": {
        "REDBIRD_API_KEY":    "your-api-key",
        "REDBIRD_PROJECT_ID": "your-project-id"
      }
    }
  }
}
Redbird MCP · Connected
Initializing Redbird MCP server...
 Authenticated · Project: acme-analytics
 Loaded 24 AI agents
 Loaded 11 workflow triggers
 Loaded 38 datasets
─────────────────────────────────────
 Ready · Waiting for tool calls

The bridge between AI conversation and real work

Your AI tools are great at thinking — planning, drafting, reasoning, summarizing. But ask them to pull your latest sales data, run a pipeline analysis, or trigger an automated report and they hit a wall. They can talk about doing things. They can't actually do them. Model Context Protocol (MCP) is the open standard that closes that gap. Redbird MCP makes Redbird's AI agents, automation workflows, and data outputs available to your AI tools through that standard — so a conversation with Claude or ChatGPT can trigger real work inside Redbird and return real results.

AI agents on demand

Your AI tool can invoke any Redbird AI agent and receive the results directly in the conversation.

Workflow execution

Trigger any Redbird workflow from your AI tool by describing what you need. Redbird runs the workflow and returns the output.

Live data and reports

Query datasets and reports produced by Redbird workflows — your AI gets structured, current data to reason about, not stale exports.

Open standard

MCP is an open protocol. Your Redbird connection works with every AI tool that adopts MCP — today and as new ones emerge.

Works with any MCP-compatible AI

Claude, ChatGPT, and any other AI tool that supports MCP can connect to Redbird through the same interface — one server, every client.

Your AI Tool
Claude, ChatGPT, or any MCP client
tool_call →
MCP
← result
Redbird MCP
Routes calls to the right Redbird capability
invoke →
Redbird
← output
Redbird Platform
AI agents, workflows, and datasets

Give your AI access to Redbird's full agent library

Redbird's AI agents handle the hardest data work — connecting to sources, cleaning and transforming data, running analysis, tagging content, matching records, and generating reports. Through Redbird MCP, your AI tools can invoke any of these agents directly from a conversation — passing inputs, triggering execution, and receiving structured outputs without any manual workflow setup.

  • Pull and analyze data from any connected source
  • Classify and tag datasets and unstructured documents at scale through tagging agents
  • Transform and join messy datasets
  • Generate reporting outputs connected to live data workflows
MCP tool call · list_agents JSON · MCP
// Your AI calls the Redbird MCP tool
{
  "tool": "redbird_list_agents",
  "arguments": {
    "category": "all"
  }
}
Tool result
[
  {
    "id":          1,
    "name":        "Connector Agent",
    "category":    "data-collection",
    "description": "Pull from any connected source"
  },
  {
    "id":          7,
    "name":        "Autotagger Agent",
    "category":    "classification",
    "description": "Classify and tag records at scale"
  },
  {
    "id":          14,
    "name":        "Analysis & Insights Agent",
    "category":    "analysis",
    "description": "Run analytics via natural language"
  },
    ...
]
Claude · with Redbird MCP MCP · Active
You
Run our Q2 marketing performance report and flag anything that moved more than 15% vs. last quarter.
AI
On it. Triggering the Q2 marketing workflow now.
tool call
redbird_run_workflow
workflow: "Q2 Marketing Performance" · variance_threshold: 0.15
redbird · workflow complete
✓ Report generated · 3 anomalies flagged
The report is done. Here's what moved more than 15%: Paid Search CTR dropped 22% vs. Q1 (2.1% → 1.6%). Email open rate rose 19% (24% → 28.5%). LinkedIn CPL increased 31% ($48 → $63). Everything else was within normal range. Want me to pull the underlying data for any of these?

Your AI can trigger real Redbird workflows — not just describe them

Every automation workflow built in Redbird — from a simple data pull to a multi-step pipeline that collects, transforms, analyzes, and reports — can be triggered through Redbird MCP. Your AI tool receives the workflow output and can reason about the results, ask follow-up questions, or trigger the next step.

  • Trigger any Redbird workflow by describing the task to your AI tool in natural language
  • Receive workflow outputs — structured data, reports, analysis results — back in the AI conversation
  • Chain workflow execution with AI reasoning — the AI can interpret results and decide what to run next
  • AI tools can check workflow run status, retrieve logs, and surface errors — full visibility into what ran

Current data and reports — available to your AI, on demand

The outputs of Redbird workflows — processed datasets, analysis results, reports — can be queried through Redbird MCP and returned to your AI tool as structured context. Your AI doesn't have to guess or work from stale information — it gets current, workflow-produced data it can reason about accurately.

  • Query any dataset produced by a Redbird workflow — filtered, sliced, and structured for AI consumption
  • Retrieve the latest report output without re-running the workflow — Redbird serves the most recent result
  • AI tools can ask follow-up questions about data — drilling into segments, comparing time periods, or surfacing anomalies
  • Permissioned access — data is served according to your Redbird access controls, so AI tools only see what the user is authorized to access
Claude · with Redbird MCP MCP · Active
You
What does our churn data look like for enterprise accounts this quarter?
AI
Let me query that dataset now.
tool call
redbird_query_dataset
dataset: "Customer Churn" · segment: "enterprise" · period: "Q2 2025"
redbird · 847 rows · structured
✓ Dataset returned
Enterprise churn this quarter is 2.1% — down from 2.8% in Q1. The biggest driver is improved retention in the 10–50 seat tier (churn dropped from 3.4% to 1.9%). Accounts >250 seats stayed flat at 0.8%.
Source: Redbird · Customer Churn Dataset · Updated today at 8:02 AM
Built on the open MCP standard — works with every AI tool that does

Redbird MCP is built on the open Model Context Protocol standard — the same protocol being adopted across the AI ecosystem. As more AI tools add MCP support, Redbird becomes accessible to them automatically. One connection. Every AI client that speaks MCP.

Cl
Claude
Desktop & Claude.ai
GPT
ChatGPT
OpenAI Agents SDK
Any MCP Client
Open protocol integration
AI Agents
Agentic pipelines & loops
IDE
IDE Extensions
Cursor, Windsurf & more

Built on the open MCP standard · New AI clients supported automatically

Connect your AI tools to Redbird — and let them get to work

Talk to our team to get set up and connect your AI tools to Redbird's agents, workflows, and data.