BoVerse Documentation
Everything you need to integrate BoVerse AI workflows into your applications — REST API, MCP server, and React components.
Overview
BoVerse provides three integration options depending on your use case:
REST API
Direct HTTP calls to execute and manage workflows from any language or platform.
MCP Server
Give Claude and other MCP-compatible AI assistants direct access to your workflows.
React Package
Drop-in UI components to embed workflow execution in your React app.
Production Base URL
https://python.boverse.ioREST API
The BoVerse REST API lets you execute workflows, list available workflows, and retrieve execution history. All endpoints require a Bearer token obtained from your BoVerse account settings.
Authentication
All API requests must include an Authorization header with your API key:
Authorization: Bearer YOUR_API_KEYGenerate an API key from your BoVerse account under Settings → API Keys.
Execute Workflow
Runs a workflow with the provided input and returns the result synchronously.
/api/execute-workflowRequest Body
{
"workflow_id": "wf_abc123",
"query": "Summarize the latest market trends for SaaS companies"
}Response
{
"success": true,
"final_output": "Here are the latest market trends for SaaS companies...",
"results": [
{
"step_id": "step_1",
"step_name": "Research Agent",
"output": "Found 12 relevant articles..."
},
{
"step_id": "step_2",
"step_name": "Summary Writer",
"output": "Here are the latest market trends for SaaS companies..."
}
]
}Example
curl -X POST https://python.boverse.io/api/execute-workflow \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"workflow_id": "wf_abc123",
"query": "What are the top AI trends in 2025?"
}'Poll Execution Status
After triggering a workflow execution, poll this endpoint to check whether the execution has completed, failed, or is still running.
/api/execution-status?execution_id={execution_id}Response
{
"success": true,
"status": "running | completed | failed",
"output_data": "string | null",
"citations": [],
"execution_log_id": "uuid-string | null",
"error_message": "string | null"
}Example
curl "https://www.boverse.io/api/execution-status?execution_id=YOUR_EXECUTION_ID" \
-H "Authorization: Bearer YOUR_API_KEY"Poll every 2-5 seconds until status is completed or failed. The execution_id is returned by the Execute Workflow endpoint, and is also available in the status_url field of the execute response.
List Workflows
Returns all workflows accessible to the authenticated user.
/api/workflows/listcurl https://python.boverse.io/api/workflows/list \
-H "Authorization: Bearer YOUR_API_KEY"{
"workflows": [
{
"id": "wf_abc123",
"name": "Market Research Agent",
"description": "Researches and summarizes market trends"
},
{
"id": "wf_def456",
"name": "Email Drafter",
"description": "Drafts professional emails from bullet points"
}
]
}Workflow Runs
Retrieves execution history for a specific workflow.
/api/workflows/{workflow_id}/runs?limit=10curl "https://python.boverse.io/api/workflows/wf_abc123/runs?limit=5" \
-H "Authorization: Bearer YOUR_API_KEY"{
"runs": [
{
"id": "run_xyz789",
"status": "completed",
"output": "Here are the latest SaaS trends...",
"created_at": "2025-01-15T10:30:00Z"
}
]
}MCP Server
The @boverse/mcp package provides a Model Context Protocol server, giving Claude and other MCP-compatible AI assistants direct access to your BoVerse workflows.
Installation
npm install -g @boverse/mcpClaude Desktop Setup
Add the following to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"boverse": {
"command": "npx",
"args": ["@boverse/mcp"],
"env": {
"BOVERSE_API_KEY": "your_api_key_here"
}
}
}
}Self-hosted deployments can point the server elsewhere with BOVERSE_API_URL. It must be an https URL (plain http is accepted only for localhost).
BOVERSE_API_KEYin the config above, so it stays on your machine instead of travelling through the model's context and any transcript logs. Upgrade from v0.1.x — it defaulted to a non-production host.Available Tools
| Tool | Description | Required params |
|---|---|---|
| boverse_execute_workflow | Execute a workflow and wait for the result | workflow_id, query |
| boverse_list_workflows | List all accessible workflows | — |
| boverse_get_workflow_runs | Get execution history for a workflow | workflow_id |
| boverse_get_run_status | Check status and output of a single run | execution_id |
Example Usage in Claude
Once configured, you can ask Claude to run your workflows directly:
React npm Package
The @boverse/react package provides drop-in React components and hooks for embedding BoVerse workflow execution in your web app.
Installation
npm install @boverse/reactInitialize Client
import { createBoVerseClient } from '@boverse/react';
const client = createBoVerseClient({
apiKey: 'YOUR_API_KEY',
baseUrl: 'https://python.boverse.io', // optional
});useWorkflow Hook
For custom UI, use the useWorkflow hook directly:
import { useWorkflow, createBoVerseClient } from '@boverse/react';
const client = createBoVerseClient({ apiKey: 'YOUR_API_KEY' });
function WorkflowRunner() {
const { execute, result, isLoading, error } = useWorkflow(client);
const handleRun = async () => {
await execute({
workflowId: 'wf_abc123',
query: 'Analyze Q4 2024 sales data',
});
};
return (
<div>
<button onClick={handleRun} disabled={isLoading}>
{isLoading ? 'Running...' : 'Run Workflow'}
</button>
{result && <pre>{result.final_output}</pre>}
{error && <p className="text-red-600">{error}</p>}
</div>
);
}All Exports
| Export | Type | Description |
|---|---|---|
| BoVerseWorkflow | Component | Full workflow UI with step visualization |
| BoVerseChat | Component | Chat-style workflow interface |
| useWorkflow | Hook | Execute workflows and manage state |
| useChat | Hook | Chat-style workflow interaction |
| createBoVerseClient | Function | Create an authenticated client instance |
| BoVerseClient | Class | Low-level API client |
Need help? Reach out at support@boverse.io