@boverse/mcp

MCP (Model Context Protocol) server for BoVerse workflow integration.

Read-only by design - This server can only execute workflows and read data. API key creation must be done manually in BoVerse for security.

Installation

npm install -g @boverse/mcp

Usage with Claude Desktop / Lovable

Add to your MCP configuration:

{
  "mcpServers": {
    "boverse": {
      "command": "npx",
      "args": ["@boverse/mcp"]
    }
  }
}

Available Tools

boverse_execute_workflow

Execute a BoVerse workflow and return the result.

Parameters:

• api_key (required): Your BoVerse API key
• workflow_id (required): The workflow to execute
• query (required): Input for the workflow

Example:

Execute my content generator workflow with the topic "AI in healthcare"

boverse_list_workflows

List all workflows accessible with your API key.

Parameters:

• api_key (required): Your BoVerse API key

Example:

Show me all my BoVerse workflows

boverse_get_workflow_runs

Get recent execution history for a workflow.

Parameters:

• api_key (required): Your BoVerse API key
• workflow_id (required): The workflow ID
• limit (optional): Max runs to return (default: 10)

Example:

Show me the last 5 runs of my summarizer workflow

Security

This MCP server is intentionally read-only:

• ✅ Execute workflows
• ✅ List workflows
• ✅ View execution history
• ❌ Create/delete API keys
• ❌ Modify workflows
• ❌ Access billing/account settings

API keys must be created manually at boverse.io → Workflow → Embed → API Keys.

Environment Variables

| Variable | Default | Description | |----------|---------|-------------| | BOVERSE_API_URL | Production URL | Custom API endpoint |

How It Works

1. User creates an API key manually in BoVerse
2. User adds the API key to their environment or passes it to tools
3. AI can use the MCP tools to execute workflows
4. Results are returned to the AI for further processing

License

MIT