MCP Server

What is MCP?

MCP (Model Context Protocol) is an open protocol that lets AI assistants interact with external tools and services. VideoConduit’s MCP server lets AI agents download videos, transcribe content, manage jobs, and more — all through natural language.

Instead of writing API calls manually, you describe what you want in plain English and your AI assistant handles the rest: calling the right tools, checking job status, and presenting results.

Quick Start

Claude Desktop / Claude Code

Add the following to your MCP client configuration:

{
  "mcpServers": {
    "videoconduit": {
      "type": "streamableHttp",
      "url": "https://videoconduit.com/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Cursor

In Cursor, open Settings → MCP and add a new server:

{
  "mcpServers": {
    "videoconduit": {
      "url": "https://videoconduit.com/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Python Client

from mcp.client.streamable_http import streamablehttp_client
from mcp import ClientSession

async with streamablehttp_client(
    "https://videoconduit.com/mcp",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
) as (read, write, _):
    async with ClientSession(read, write) as session:
        await session.initialize()
        result = await session.call_tool(
            "get_video_info",
            {"url": "https://youtube.com/watch?v=dQw4w9WgXcQ"},
        )
        print(result.content[0].text)

MCP Inspector (Debugging)

npx @modelcontextprotocol/inspector \
  --url https://videoconduit.com/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

Authentication

The MCP server uses the same API keys as the REST API. Pass your key as a Bearer token in the Authorization header:

Authorization: Bearer vc_your_api_key

Create API keys from the dashboard or see the Authentication docs for details. All tool calls are authenticated and scoped to your account.

Available Tools

VideoConduit exposes 20 tools organized into four categories.

Info & Status

Video Processing

Management

Convenience

Example Workflows

Download and Transcribe a YouTube Video

1. AI calls get_video_info to check the video metadata and duration
2. AI calls download_and_transcribe to start both jobs in parallel
3. AI calls get_job_status for each job to check completion
4. AI presents the download link and transcript to you

Analyze a Video’s Engagement

1. AI calls get_video_info to get metadata and engagement metrics
2. AI calls extract_comments to get audience reactions
3. AI waits for completion, then summarizes the engagement data and top comments

Manage Your Account

1. AI calls get_credit_balance to check remaining credits
2. AI calls list_jobs to see recent activity
3. AI calls create_webhook to set up notifications for job completion

Error Handling

Tools return error dictionaries instead of throwing exceptions. Your AI agent will see these errors and can explain them or retry with different parameters.

Common Errors

Insufficient credits:

{"error": "Insufficient credits. Need 3, have 1."}

Invalid URL:

{"error": "Could not process URL: Unsupported URL or video not found."}

Job not found:

{"error": "Job not found"}

Authentication failure:

Missing or invalid API keys return a 401 JSON-RPC error response before any tool is called.

Rate Limits & Billing

• MCP tool calls use the same credits as the equivalent REST API endpoint
• Tool calls are not separately rate-limited beyond the standard API key rate limits
• Check your balance anytime with get_credit_balance
• See Rate Limits for per-plan limits