# CrewAI Platform MCP for AI Agents AI Agent Connect

> CrewAI Platform MCP lets you manage and trigger multi-agent workflows directly from your AI client. You can launch autonomous crews, monitor their progress in real-time, and audit agent roles without switching back and forth between your dashboard and your chat. It's built for anyone running complex, multi-step agentic processes who needs a direct line of command.

## Overview
- **Category:** superpower
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_vQl1sjKmbDR8pDbzxB1A2J0p9pCxG6EwZ7FmwnqF/ai-agent-connect
- **Tags:** multi-agent, autonomous-agents, workflow-automation, llm-orchestration, task-execution

## Description

Imagine you've built a complex system of autonomous agents that handle everything from market research to content production. Usually, managing these means jumping between a dashboard, checking logs, and manually feeding inputs into different workflows. This Connector changes that by bringing the entire CrewAI Platform into your primary workspace. Instead of hunting for specific run IDs or squinting at status logs, you can just tell your agent to start a specific crew or ask for a status update on a running task. It handles the heavy lifting of fetching the right JSON payloads and managing the execution boundaries. You get to stay in your flow while your agents do the work. Because Vinkius makes it so easy to connect these types of specialized tools, you can move from a high-level idea to a running multi-agent execution in seconds. It's about having a single point of control for your entire autonomous fleet.

## Tools

### get_status
See the current state of an active execution. It helps you track where an agent is stuck or when a task is finished.

### kickoff_crew
Start a new multi-agent run using a JSON payload. This is the primary way to put your autonomous workflows into motion.

### get_inputs
Find the exact data points required to start a specific crew. Use this to ensure your agent provides the right information before a run begins.

## Prompt Examples

**Prompt:** 
```
What crews do I have available right now?
```

**Response:** 
```
I found the following crews in your account:

| Crew Name | ID | Status |
| :--- | :--- | :--- |
| Market Research Team | `crew_abc` | Active |
| Content Creation Squad | `crew_def` | Ready |
| Data Analysis Group | `crew_ghi` | Ready |

Which one would you like to inspect or start?
```

**Prompt:** 
```
Start the Market Research crew with the topic 'EV Battery Tech'.
```

**Response:** 
```
I've kicked off the **Market Research Team** (`crew_abc`).

**Run Details:**
*   **Run ID:** `run_9921`
*   **Topic:** EV Battery Tech
*   **Status:** Agents are currently initializing the research phase.

I'll monitor the progress for you. Just ask for a status update anytime.
```

**Prompt:** 
```
Tell me what the 'Senior Analyst' agent is supposed to do.
```

**Response:** 
```
Here are the details for the **Senior Analyst** agent:

*   **Role:** Senior Research Analyst
*   **Backstory:** Expert in gathering and synthesizing complex data from multiple sources to provide actionable insights.
*   **Primary Goal:** Identify key market trends and summarize them into executive briefs.

Would you like to see the specific tasks assigned to this agent?
```

## Capabilities

### Kick off autonomous crews
Start complex multi-agent workflows with specific data inputs.

### Monitor live execution
See exactly where your agents are in a sequential or parallel task.

### Audit agent backstories
Review the roles and goals of every agent in your deployed workflows.

### Interrupt active runs
Send an immediate stop signal to any running crew to prevent unnecessary costs.

### Inspect webhook criteria
Check the validation rules for how your crews post their final results.

### Map workflow topology
Get a complete JSON blueprint of your agent graph structure.

## Use Cases

### Market Research
A researcher asks the agent to start a Competitor Analysis crew. The agent uses kickoff_crew to trigger the run and tells the user when it's done.

### Content Pipeline
A social media manager wants to see the status of a Blog Poster crew. The agent calls get_status and summarizes the current step.

### Debugging
A developer wants to see if an agent is following its instructions. They ask the agent to pull the backstory for a specific role to verify the logic.

### Workflow Setup
A product lead wants to know what data they need to provide for a new Lead Gen crew. The agent uses get_inputs to list the required fields.

## Benefits

- Stop manual dashboard switching by using kickoff_crew to start runs directly from your chat.
- Reduce wasted compute by using get_status to monitor progress and get_inputs to ensure data accuracy.
- Improve agent reliability by auditing backstories and role-playing parameters through the Connector.
- Streamline production by inspecting webhook validation criteria to ensure data lands where it should.
- Get a clear bird's-eye view of your entire agent graph topology using the JSON blueprint export.

## How It Works

The bottom line is you get a direct command line for your autonomous agent workflows inside your favorite chat or IDE.

1. Subscribe to the CrewAI Platform MCP on Vinkius.
2. Paste your Agent Token from the AMP Dashboard Status tab.
3. Use your AI client to trigger crews, check statuses, or audit agent roles.

## Frequently Asked Questions

**How can I use the CrewAI Platform MCP to manage my agents?**
You can use it to trigger specific workflows, check the live status of running tasks, and audit agent roles directly from your AI chat interface.

**Can I start a CrewAI workflow using just a text prompt?**
Yes, you can tell your agent to start a specific crew with a natural language prompt, and it will handle the JSON payload for you.

**How do I see the live status of a running multi-agent task?**
You can ask your agent for a status update on any active run. It will retrieve the current execution state and tell you what the agents are doing right now.

**Can this Connector help me debug my agent backstories?**
Absolutely. You can ask the agent to pull the backstory and role definitions for any agent in your fleet to ensure they are following your logic.

**Is it possible to stop a running CrewAI crew through the chat?**
Yes, you can send an instant interrupt signal to hard-stop any active run if you need to stop a process immediately.

**How does the CrewAI Platform MCP handle input requirements?**
It can identify the exact bounded inputs required to kickoff a specific crew, so you know exactly what data to provide before starting a run.

**Can my agent kickoff a new CrewAI workflow?**
Yes. Use the 'kickoff_crew' tool. Provide the Crew ID and a JSON object with the required inputs. The agent will activate the multi-agent processing immediately, returning a run ID for tracking.

**How do I monitor the progress of an active agent run?**
Use the 'get_run_status' tool with your Crew ID and Run ID. Your agent will grab the live execution state, showing you which agents are currently working and which tasks have been completed.

**Can I cancel a running crew via the agent?**
Absolutely. The 'cancel_run' tool dispatches an instant interrupt signal to the CrewAI platform, hard-stopping active LLM contexts and terminating the execution flow immediately.