Agent automation

Video encoding MCP server for agents and workflow tools

Connect Convertrilo to MCP clients so agents can create encoding jobs, check status, wait for completion, cancel work, and inspect token balance.

Convertrilo MCP gives agents a narrow stdio tool surface over the same production API. It is built for explicit source URLs, scoped API keys, idempotent workflows, and terminal job states.

Read MCP setupOpen npm packageDeveloper settings
MCP client configautomation
npm install -g @convertrilo/sdk
convertrilo login

{
  "mcpServers": {
    "convertrilo": {
      "command": "convertrilo-mcp"
    }
  }
}

Let agents submit encode jobs without exposing broad dashboard access.

Read job status and token balance from MCP clients.

Wait for success, failure, or cancellation before taking follow-up actions.

Use the same saved credentials as the Convertrilo CLI.

Workflow

Built for jobs that should run without a human babysitting a tab

Each automation page maps to the same production path: API request, durable job row, queue, Rust worker, storage upload, terminal event, and reconciliation.

  1. Step 1

    Install @convertrilo/sdk globally or provide an absolute path to the local MCP build.

  2. Step 2

    Run convertrilo login to save a scoped API key from Developer Settings.

  3. Step 3

    Add convertrilo-mcp to your MCP client configuration.

  4. Step 4

    Ask the client to list tools, then call create_encode_job, get_job_status, wait_for_job, cancel_job, or get_token_balance.

Small tool surface

The server exposes focused encode, status, wait, cancel, and balance tools instead of raw arbitrary account access.

stdio transport

Run the MCP server locally in clients that support command-based stdio servers. No hosted MCP URL is required.

Shared CLI credentials

MCP reads CONVERTRILO_API_KEY when provided, or the key saved by convertrilo login in ~/.convertrilo/config.json.

Automation questions

Which MCP clients can use Convertrilo?

Any MCP client that can launch a local stdio server command can run convertrilo-mcp with a scoped API key.

Does the MCP server need a public URL?

No. The current MCP server is stdio-based and runs on the user's machine or agent host.

Can agents create paid priority jobs?

Yes, if your API key has the required job scopes and your workflow allows priority. Use narrow keys and explicit tool instructions.

Related automation pages

API integration guideSDK, idempotency, S3, Google Drive, and webhook workflows.Video encoding APICreate video encoding jobs from your backend with API keys, idempotency, metadata, priority queues, webhooks, and storage destinations you control.Video encoding CLIAutomate video compression from terminals, deployment jobs, content pipelines, and scheduled scripts with a Convertrilo CLI workflow.Video encoding webhooksReceive signed job.completed, job.failed, and job.canceled events so your app can publish assets, update records, and notify customers automatically.Batch video encodingQueue many videos from folders, asset libraries, imports, and customer batches with metadata, idempotency, webhooks, and storage destinations.Automated S3 video compressionCompress videos from S3-compatible storage and write finished MP4 outputs back to your own bucket with API keys, idempotency, and worker execution reports.Agent video workflow automationGive MCP-capable agents a narrow video encoding tool surface for creating jobs, waiting for completion, inspecting reports, and handing outputs back to your workflow.Folder ingest automationQueue one encode job per video discovered in S3-compatible or Google Drive folders, with caps, metadata, output destinations, and webhook reconciliation.

Start with a real estimate

Choose your codec, resolution, FPS, bitrate, quality, and optimization settings. Convertrilo shows the NEU estimate before credits are reserved, and failed jobs release reserved credits back to your balance.

Start FreeCalculate Pricing