Skip to main content

AI Agent Integration via MCP

Connect AI agents to Prometheux using the Model Context Protocol (MCP) — an open standard that enables AI assistants to interact with external tools and data sources.

Prometheux offers three MCP integration options:

  • Local MCP Server (prometheux-mcp): Runs on your machine — works with Claude Desktop, Claude Code, Cursor, and any stdio-capable client. Recommended for coding agents.
  • Remote MCP Server (px-remote-mcp-server): Cloud-hosted option that works with any MCP-compatible client via OAuth authentication
  • Built-in MCP Server (Snowflake Native App only): Automatically registered when the app starts — Cortex Agents and Snowflake Intelligence can discover Prometheux tools with no external setup. See the Native App guide for details.

What is MCP?​

The Model Context Protocol allows AI agents to:

  • Discover available tools and resources
  • Execute operations through a standardized interface
  • Access external data sources and APIs

With Prometheux's MCP integration, you can use natural language to interact with your ontologies, list concepts, and execute reasoning — all directly from your preferred AI agent.

Which Option Should I Use?​

Best for: Coding agents (Claude Code, Cursor), Claude Desktop, development environments, on-premise deployments

Pros:

  • Simple pip install
  • Runs locally (no external dependencies)
  • Full control over credentials — no OAuth flow required
  • Works with any Prometheux instance (cloud or on-premise)
  • Compatible with Claude Desktop, Claude Code, Cursor, and any stdio-capable client

Cons:

  • Needs to be installed on each machine

Use this when:

  • You're using a coding agent (Claude Code, Cursor)
  • You're using Claude Desktop
  • You want a simple, local installation
  • You're developing or testing locally

Get Started with Local MCP →

Supported MCP Clients​

We provide step-by-step connection guides for the following clients:

ClientStatusGuide
Cursor✅ SupportedConnect Cursor →
Claude Code (Terminal CLI)✅ SupportedConnect Claude Code →

Chat & Enterprise Clients (Remote Server)​

ClientStatusGuide
Claude (Desktop & Web)✅ SupportedConnect Claude →
ChatGPT (OpenAI)✅ SupportedConnect ChatGPT →
Genie Code (Databricks)✅ SupportedConnect Genie Code →
Snowflake Cortex (Cortex Code & Intelligence)✅ SupportedConnect Snowflake Cortex → (Native App users: built-in MCP)

Don't see your client? For stdio-capable clients (IDEs, terminal agents), use the Local MCP Server. For browser-based clients, the Remote MCP Server exposes a standard OAuth endpoint.

Available Tools​

Both MCP servers dynamically expose the following tools to AI agents. The tool catalog is served by the Prometheux backend, so new capabilities are available automatically as they are added to the platform.

Projects & Data Sources​

ToolDescriptionKey Parameters
list_projectsList all projects the user has access toscope
create_projectCreate a new empty projectname, description, scope
list_data_sourcesList all data sources with their predicate names, bind annotations, field schemas, and row countsscope

Concepts​

ToolDescriptionKey Parameters
list_conceptsList all concepts available in a projectproject_id, scope
get_conceptGet full detail for a single concept including its Vadalog codeproject_id, concept_name, scope
create_conceptSave a Vadalog program as a new named conceptproject_id, vadalog_code, description, scope
update_conceptOverwrite the Vadalog code of an existing conceptproject_id, concept_name, vadalog_code, description, scope
run_conceptExecute a concept to derive new knowledge through reasoning (supports long-running executions with progress notifications)project_id, concept_name, params, scope, force_rerun, persist_outputs

Vadalog​

ToolDescriptionKey Parameters
search_vadalog_docsSearch the Vadalog documentation and return ready-to-use code examplesquery
generate_vadalogGenerate a Vadalog program from a natural language descriptiondescription, predicates_and_schemas
fix_vadalogFix a Vadalog program that failed to compile or executevadalog_code, error_message, original_request
generate_sample_factsGenerate realistic sample input or output facts for a set of predicatespredicates_and_schemas, domain_context, vadalog_rules, fact_type

Ontology​

ToolDescriptionKey Parameters
get_ontologyLoad the current ontology graph for a projectproject_id, scope
save_ontologySave a complete ontology graph for a projectproject_id, ontology_data, scope
set_concept_ontology_roleAssign or update a single concept's role in the ontology graphproject_id, concept_name, ontology_type, edge_source, edge_target, scope

Dashboards​

ToolDescriptionKey Parameters
list_dashboardsList all dashboards across every project in the workspacescope
get_dashboardLoad a single dashboard with its full definitionproject_id, dashboard_id, scope
save_dashboardCreate or update a dashboard for a projectproject_id, dashboard, scope
delete_dashboardPermanently delete a dashboardproject_id, dashboard_id, scope

Company Knowledge​

ToolDescriptionKey Parameters
get_company_infoSearch the Prometheux company knowledge basequery
Parameter Defaults

Most tools accept a scope parameter ("user" or "organization", default "user"). Boolean parameters like force_rerun and persist_outputs default to true and false respectively.

Example Usage​

Once configured, just chat naturally with your AI agent. The agent will automatically use the Prometheux MCP tools when relevant.

Example queries:

"What projects do I have?"

"What concepts are available in my customer-analytics project?"

"Run the churn_prediction concept in the customer-analytics project"

"Generate a Vadalog program that finds all transitive suppliers for a company"

"Create a dashboard showing the results of churn_prediction"

"What does the Vadalog documentation say about recursive rules?"

The AI agent will automatically:

  • Call the appropriate Prometheux MCP tools
  • Parse the results
  • Present them in a readable format
  • Answer follow-up questions about the data

Learn More​