Purpose
cognigy-ai-mcp-management-server is a local MCP server used to manage Cognigy.AI virtual agents through Cognigy.AI's Management API. It enables AI coding assistants (such as Claude, Cursor, etc.) to programmatically build, configure, test, and operate Cognigy.AI conversational AI agents, thereby replacing manual click operations in the Cognigy UI.
This server is an independent open-source project and is not affiliated with, endorsed by, or sponsored by Cognigy or NiCE. Using it requires your own valid Cognigy.AI account and API Key, and compliance with Cognigy's commercial license terms. "Cognigy" and "Cognigy.AI" are trademarks of their respective owners.
Target users: developers, solution architects, and system integration (SI) partners developing on Cognigy.AI.
Main Capabilities
Through natural language, you can:
- Create and manage flows, nodes, and intents
- Automatically run NLU training and regression tests
- Deploy snapshots and packages across environments
- Configure LLM, Knowledge AI, and integrations
The server provides a total of 132 MCP tools, organized by domain. Its security and operational features include:
- Secure by default: all mutating tools use dryRun: true by default
- Async-aware: long-running operations are polled until completion
- Secure: API Keys are kept only in memory, and secrets are automatically redacted
- Pagination: all list operations support limit and skip
Usage
Prerequisites
- Node.js 20.0.0 or higher
- A Cognigy.AI account with API access (Trial, SaaS, or on-premises deployment)
- An API Key generated from the Cognigy UI (My Profile → API Keys)
Important: The @cognigy/rest-api-client package is licensed under the "Cognigy Proprietary License". By installing this package, you agree to Cognigy's license terms.
Configure Environment Variables
Environment variables must be set before running, including common Base URL, etc. (for the specific variable names and values, please refer to the project documentation).
MCP Client Configuration
- macOS: add to ~/Library/Application Support/Claude/claude_desktop_config.json
- Windows: add to %APPDATA%\Claude\claude_desktop_config.json
- Add in .mcp.json in the project root directory
- Add in .cursor/mcp.json
Installation and Dependencies
This package requires @cognigy/rest-api-client as a peer dependency, which means:
- You must install it separately
- Installing Cognigy's package means you accept Cognigy's license terms
- This project does not bundle or redistribute any Cognigy code
You are responsible for:
- Reviewing and accepting Cognigy's proprietary license terms
- Having the appropriate rights to use the Cognigy API
- Ensuring API credentials are secure
The Cognigy REST API client is the official SDK maintained by Cognigy GmbH, published on npm under the "Cognigy Proprietary License".
Development and Testing
The project includes live API tests and an optional type regeneration step (when generated types need to be updated from an updated Cognigy API). This MCP server is released under the MIT License, see LICENSE for details. Note: The MIT License applies only to this MCP server code; dependencies (especially @cognigy/rest-api-client) are subject to their respective licenses.
Contribution guidelines: Fork the repository, create a feature branch (git checkout -b feature/amazing-feature), follow existing code patterns (Zod schemas, dryRun flags, etc.), add tests for new tools, run npm test && npm run build before committing, then submit a Pull Request.
Development guidelines:All tools must have Zod input validation; change-type tools must support a dryRun flag (default true); tool descriptions should be easy for an LLM to understand (1-2 sentences); never log or expose API Keys/secrets; test against a mock server first before connecting to the live API.
Applicable Scenarios
- Developers, solution architects, and SI partners building conversational AI agents on Cognigy.AI
- Want to use natural language instead of UI clicks to manage flows, nodes, and intents
- Need to automatically run NLU training and regression tests
- Need to deploy snapshots and packages across environments
- Need to configure LLM, Knowledge AI, and integrations
Limitations
- Requires your own valid Cognigy.AI account and API Key
- Not an official Cognigy/NiCE project or affiliated with them, and no official support is provided
- Use is subject to Cognigy's commercial license terms, and @cognigy/rest-api-client is under a proprietary license
- This server's code is MIT licensed, but dependencies are subject to their respective licenses
- The related materials do not provide specific pricing, cost, or performance data
