PX APIs

The Prometheux platform provides multiple ways to interact with all platform features programmatically:

Python SDK (prometheux_chain): A high-level Python library for seamless integration
REST API: Direct HTTP endpoints for maximum flexibility and language-agnostic access
MCP Integration: Connect AI agents to your ontologies via Model Context Protocol
- Local MCP Server (prometheux-mcp): For Claude Desktop with local installation
- Remote MCP Server (px-remote-mcp-server): For any MCP client (Claude, Genie Code, and more) with OAuth

All interfaces provide equivalent functionality—choose the one that best fits your workflow.

Quick Start

Python SDK
MCP (Claude Desktop)
MCP (Remote)
REST API

Installation

pip install --upgrade prometheux_chain

Configuration

import prometheux_chain as px
import os

# Set your authentication token
os.environ['PMTX_TOKEN'] = 'YOUR_PMTX_TOKEN'

# Configure the backend URL
px.config.set('JARVISPY_URL', "https://api.prometheux.ai/jarvispy/{organization}/{username}")

First API Call

# Create a new project
ontology_id = px.save_ontology(ontology_name="my_first_ontology")
print(f"Created project: {project_id}")

Install into Jupyter

To manually install the library in your Jupyter Lab or Jupyter Notebook follow these steps:

Create a new notebook with kernel Python and run the following commands in a dedicated cell

Clone the repo

!git clone https://github.com/prometheuxresearch/prometheux_chain.git

Navigate into the project and install it

!cd prometheux_chain && pip install -e .

Copy the project into site-packages to allow importing and using it

!cd prometheux_chain && cp -r prometheux_chain /opt/app-root/lib/python3.x/site-packages

Restart the kernel

To update an existing installation of Prometheux Chain follow these steps:

Uninstall the project
!pip uninstall prometheux_chain -y

Clone the repo if it is not present

!git clone https://github.com/prometheuxresearch/prometheux_chain.git

else update it to the latest changes

!cd prometheux_chain && git pull

Navigate into the project and install it

!cd prometheux_chain && pip install -e .

Remove an existing installation if present

!rm -r /opt/app-root/lib/python3.x/site-packages/prometheux_chain

Copy the project into site-packages to allow importing and using it

!cd prometheux_chain && cp -r prometheux_chain /opt/app-root/lib/python3.x/site-packages

Restart the kernel

Automated Installation (Recommended)

macOS/Linux:

curl -sSL https://raw.githubusercontent.com/prometheuxresearch/px-mcp-server/main/install.sh -o install.sh
chmod +x install.sh
./install.sh

Windows (PowerShell):

Invoke-WebRequest -Uri "https://raw.githubusercontent.com/prometheuxresearch/px-mcp-server/main/install.ps1" -OutFile "install.ps1"
.\install.ps1

The script will install everything and configure Claude Desktop automatically.

Manual Installation

pipx install prometheux-mcp

Then add to Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "prometheux": {
      "command": "/Users/YOUR_USERNAME/.local/bin/prometheux-mcp",
      "args": ["--url", "https://api.prometheux.ai"],
      "env": {
        "PROMETHEUX_TOKEN": "your_token",
        "PROMETHEUX_USERNAME": "your_username",
        "PROMETHEUX_ORGANIZATION": "your_org"
      }
    }
  }
}

Run which prometheux-mcp to find the exact path on your system.

First Interaction

Restart Claude Desktop and ask:

“What concepts are available in my project-name project?”

See the full MCP documentation for detailed setup and usage.

Cloud-Hosted with OAuth

The remote MCP server works with any MCP-compatible client — Claude, Genie Code, and more — using OAuth for authentication.

Server URL

https://api.prometheux.ai/px-remote-mcp-server

Connect Your Client

We provide step-by-step guides for specific clients:

Claude (Desktop & Web) — Add as a custom connector in Claude settings
Genie Code (Databricks) — Register as an external MCP server

For other MCP clients, see the Remote MCP Server page for OAuth endpoints and connection parameters.

First Interaction

Once connected, ask your AI agent:

“What concepts are available in my project-name project?”

The remote server handles authentication and communicates with your Prometheux instance.See the full MCP documentation for detailed setup and troubleshooting.

Base URL

All API endpoints are relative to your Prometheux platform instance:

https://api.prometheux.ai/jarvispy/{organization}/{username}/api/v1/

Where:

{organization} is your organization identifier
{username} is your username

Authentication

All API requests require authentication using a Bearer token:

Authorization: Bearer YOUR_JWT_TOKEN

First API Call

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/ontologies/save" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "project": {
      "name": "my_first_project"
    }
  }'

Response Format

All REST API responses follow a consistent format:

{
  "data": {},           // Response data (varies by endpoint)
  "message": "string",  // Human-readable message
  "status": "success"   // Status: "success" or "error"
}

The Python SDK automatically handles responses and returns the relevant data directly.

API Categories

The API is organized into the following categories:

Category	Description
REST API conventions	Base URL, response envelope, scoping, pagination, errors
Authentication	Obtain, list, and revoke API tokens
Ontologies	Ontology management, export/import, snapshots
Concepts	Define, run, and read results from concepts
Data Sources	Connect data sources, infer schema, upload and preview data
Knowledge Graph	Build graphs and run graph analytics
Vadalog	Analyze and evaluate Vadalog programs
Vadalingo	NL, SQL, RDF, and OWL to Vadalog translation
Apps	App CRUD and versioning
Schedules	Automate concept runs
Context Layer	Manage and search the agent knowledge base
Config	SDK configuration and LLM setup

Error Handling

Python SDK
REST API

The SDK raises exceptions for errors:

import prometheux_chain as px

try:
    result = px.run_concept(project_id="invalid_id", concept_name="test")
except Exception as e:
    print(f"Error: {e}")

The REST API uses standard HTTP status codes:

Code	Description
`200`	Success
`400`	Bad Request (invalid parameters)
`401`	Unauthorized (invalid or missing token)
`500`	Internal Server Error

Error responses include detailed messages in the response body:

{
  "data": null,
  "message": "Project not found",
  "status": "error"
}

Introduction

Get Started

Platform

Integrations

Examples

Vadalog Reference

Resources

Quick Start

Installation

Configuration

First API Call

Install into Jupyter

Automated Installation (Recommended)

Manual Installation

First Interaction

Cloud-Hosted with OAuth

Server URL

Connect Your Client

First Interaction

Base URL

Authentication

First API Call

Response Format

API Categories

Error Handling

Requirements

Python SDK

REST API

​Quick Start

​Installation

​Configuration

​First API Call

​Install into Jupyter

​Automated Installation (Recommended)

​Manual Installation

​First Interaction

​Cloud-Hosted with OAuth

​Server URL

​Connect Your Client

​First Interaction

​Base URL

​Authentication

​First API Call

​Response Format

​API Categories

​Error Handling

​Requirements

​Python SDK

​REST API

Quick Start

Installation

Configuration

First API Call

Install into Jupyter

Automated Installation (Recommended)

Manual Installation

First Interaction

Cloud-Hosted with OAuth

Server URL

Connect Your Client

First Interaction

Base URL

Authentication

First API Call

Response Format

API Categories

Error Handling

Requirements

Python SDK

REST API