cognigy-ai-mcp-management-server
This server enables programmatic management of Cognigy.AI virtual agents, allowing AI assistants to build, configure, test, and operate agents via natural language or API calls through 132 MCP tools.
Project & Flow Management: List projects, list/get flows, get flow settings, and retrieve latest log entries.
Node Operations: List, get, search, create, update, delete, and move nodes; generate AI-powered node content (e.g., Say messages or Adaptive Cards).
Intent & NLU: CRUD intents, manage example sentences, generate new sentences via AI, train NLU models, score utterances, audit NLU quality, run regression tests.
Endpoint & Session Management: List/get endpoints, inject/reset session context, fetch conversations and transcripts.
Testing & Playbooks: List/get/run playbooks, review runs, execute regression suites.
Snapshots & Packages: Create, delete, restore, package, download, upload, diff, and promote snapshots; create, merge, upload, download, delete packages; clone flows.
Connections & LLMs: Manage API connections (secrets redacted), configure and test LLM providers, clone/delete LLMs.
NLU Connectors: Integrate external NLU services (Dialogflow, LUIS, Watson, Alexa, Lex, custom code).
Knowledge AI (RAG): Manage knowledge stores, sources, chunks, connectors for RAG; create/update/delete/run connectors.
Functions & Extensions: List/get custom code functions; upload/configure/manage extensions.
Contact Profiles & Analytics: Manage contact profiles (CRUD, merge, export for GDPR) and access analytics metrics.
Audit & Search: Retrieve audit event logs and perform organization-wide resource search.
Async Task Monitoring: List and get status of long-running tasks (training, snapshot creation, etc.) with polling support.
Safety: All mutating tools default to
dryRun: trueto validate operations before applying changes; secrets are redacted in responses.
Allows configuration and management of external NLU connectors such as Dialogflow for natural language understanding in Cognigy.AI virtual agents.
Cognigy.AI Management MCP Server

Model Context Protocol server for managing Cognigy.AI virtual agents through the Management API
This is an independent, open-source MCP server and is not affiliated with, endorsed by, or sponsored by Cognigy or NiCE. It requires your own valid Cognigy.AI account and API key, used in accordance with Cognigy's commercial license terms. "Cognigy" and "Cognigy.AI" are trademarks of their respective owners.
What is this?
cognigy-ai-mcp-management-server is a local MCP server that enables AI coding assistants (Claude, Cursor, etc.) to build, configure, test, and operate Cognigy.AI conversational AI agents programmatically.
Instead of clicking through the Cognigy UI, you can now:
Create and manage flows, nodes, and intents via natural language
Run NLU training and regression tests automatically
Deploy snapshots and packages across environments
Configure LLMs, Knowledge AI, and integrations
Built for: Developers, solution architects, and SI partners who build on Cognigy.AI.
Related MCP server: MCP Starter
Requirements
Node.js 20.0.0 or higher
Cognigy.AI account with API access (Trial, SaaS, or on-premises)
API Key generated from Cognigy UI (My Profile → API Keys)
Installation
# 1. Install the Cognigy REST API client (required peer dependency)
npm install @cognigy/rest-api-client
# 2. Install this MCP server
npm install -g cognigy-ai-mcp-management-serverImportant: The @cognigy/rest-api-client package is licensed under "Cognigy Proprietary License". By installing it, you agree to Cognigy's license terms.
Configuration
Set environment variables before running:
# Required: Your Cognigy API endpoint
export COGNIGY_BASE_URL=https://api-trial.cognigy.ai
# Required: Your API key (never commit this!)
export COGNIGY_API_KEY=your-api-key-here
# Optional: Default project for operations
export COGNIGY_DEFAULT_PROJECT_ID=your-project-idCommon Base URLs:
Environment | Base URL |
Trial |
|
SaaS EU |
|
SaaS US |
|
Dedicated |
|
MCP Client Configuration
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"cognigy": {
"command": "npx",
"args": ["cognigy-ai-mcp-management-server"],
"env": {
"COGNIGY_BASE_URL": "https://api-trial.cognigy.ai",
"COGNIGY_API_KEY": "your-api-key-here"
}
}
}
}Claude Code
Add to .mcp.json in your project root:
{
"mcpServers": {
"cognigy": {
"command": "npx",
"args": ["cognigy-ai-mcp-management-server"],
"env": {
"COGNIGY_BASE_URL": "https://api-trial.cognigy.ai",
"COGNIGY_API_KEY": "your-api-key-here"
}
}
}
}Cursor
Add to .cursor/mcp.json:
{
"mcpServers": {
"cognigy": {
"command": "npx",
"args": ["cognigy-ai-mcp-management-server"],
"env": {
"COGNIGY_BASE_URL": "https://api-trial.cognigy.ai",
"COGNIGY_API_KEY": "your-api-key-here"
}
}
}
}Available Tools
This server provides 132 MCP tools organized by domain:
Category | Tools | Description |
Projects & Flows | 5 | List projects, flows, flow settings, logs |
Nodes | 9 | CRUD operations, search, move, AI output generation |
Intents & NLU | 10 | Intent management, training, scoring, audit |
Endpoints | 2 | List and inspect endpoint configurations |
Sessions | 2 | Inject/reset conversation context |
Conversations | 3 | Fetch conversations, transcripts |
Playbooks & Testing | 6 | Run playbooks, regression tests |
Snapshots | 9 | Create, restore, package, diff, promote |
Packages | 7 | Create, merge, upload, download |
Connections | 5 | Manage API connections (secrets redacted) |
LLMs | 7 | Configure generative AI providers |
NLU Connectors | 5 | External NLU (Dialogflow, LUIS, etc.) |
Knowledge AI | 21 | Stores, sources, chunks, connectors (RAG) |
Functions | 9 | Custom code functions, instances |
Extensions | 6 | Upload, configure, manage extensions |
Contact Profiles | 11 | Profile CRUD, merge, export (GDPR) |
Analytics | 4 | Conversation, call, knowledge metrics |
Audit | 2 | Audit event logs |
Handover | 7 | Live agent providers and services |
Search | 1 | Organization-wide resource search |
Tasks | 2 | Async task status tracking |
Key Features
Safe by default: All mutating tools use
dryRun: trueby defaultAsync-aware: Long-running operations poll until completion
Security: API keys only in memory, secrets automatically redacted
Pagination: All list operations support
limitandskip
Development & Testing
Live API Testing
# Create .env from template
cp .env.example .env
# Edit .env with your credentials
# Run development server
npm run devBuild
npm run build # Compile TypeScript
npm test # Run test suite (49 tests)
npm run lint # Check code styleRegenerate Types (optional)
If you need to update the generated types from a newer Cognigy API:
npm run update:spec # Download latest OpenAPI spec
npm run gen:types # Regenerate TypeScript typesImportant: Dependencies
This package requires @cognigy/rest-api-client as a peer dependency. This means:
You must install it separately (see Installation above)
You accept Cognigy's license terms by installing their package
We do not bundle or redistribute any Cognigy code
You are responsible for:
Reviewing and accepting Cognigy's proprietary license terms
Having appropriate rights to use the Cognigy API
Keeping your API credentials secure
The Cognigy REST API client is the official SDK maintained by Cognigy GmbH and is published on npm under "Cognigy Proprietary License".
License
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 (particularly @cognigy/rest-api-client) are subject to their own licenses.
Contributing
Contributions are welcome! Please:
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 buildbefore committingSubmit a Pull Request
Development Guidelines
All tools must have Zod input validation
Mutating tools must support
dryRunflag (default: true)Tool descriptions should be LLM-friendly (1-2 sentences)
Never log or expose API keys/secrets
Test against mock server before live API
Support
Issues: GitHub Issues
Cognigy Docs: docs.cognigy.com
MCP Spec: modelcontextprotocol.io
Built with Model Context Protocol SDK
Available Tools
138 toolsaudit_nluARead-onlyIdempotent
Audits Cognigy.AI NLU quality for a flow. Identifies intents with too few training sentences, disabled intents, and optionally checks for overlapping intents. Use this before deployment to ensure NLU quality.
Inputs: flowId (required), projectId, checkOverlap, minSentences.
clone_flowA
Clones a Cognigy.AI flow within the same project. Creates an exact copy of the flow including all nodes, intents, and configurations. The cloned flow gets an auto-generated name. MUTATING: Set dryRun=false to clone.
Inputs: flowId (required), dryRun.
clone_llmA
Clones a Cognigy.AI large language model configuration. Creates a copy with the same settings that can be modified independently. MUTATING: Set dryRun=false to clone.
Inputs: largeLanguageModelId (required), dryRun.
create_connectionA
Creates a new Cognigy.AI connection for external service integration. Connections securely store credentials like API keys, passwords, and tokens. MUTATING: Set dryRun=false to create.
Inputs: name (required), type (required), extension (required), projectId (required), +2 more.
create_contact_profileA
Creates a new Cognigy.AI contact profile. Contact profiles persist user data across sessions for personalization. MUTATING: Set dryRun=false to create.
Inputs: contactId (required), projectId (required), dryRun, profile, +1 more.
create_functionA
Creates a new Cognigy.AI Function. Functions are custom code modules for integrations, scheduled jobs, or computations. MUTATING: Set dryRun=false to create.
Inputs: name (required), projectId (required), code, dryRun.
create_handover_providerA
Creates a new Cognigy.AI handover provider for live agent escalation, based on one of the services from list_handover_services. The provider starts with the service's default properties; change them with update_handover_provider. MUTATING: Set dryRun=false to create.
Inputs: name (required), type (required), projectId (required), dryRun.
create_intentA
Creates a new intent in a Cognigy.AI flow for NLU recognition. MUTATING: This modifies the flow. Use dryRun=true (default) to validate first. After creating, use train_intents to train the NLU model.
Inputs: name (required), flowId (required), tags, rules, +7 more.
create_knowledge_chunkA
Creates a new Cognigy.AI knowledge chunk manually. Use this to add specific text segments that should be searchable via RAG. The chunk will be embedded automatically. MUTATING: Set dryRun=false to create.
Inputs: text (required), sourceId (required), knowledgeStoreId (required), title, +2 more.
create_knowledge_connectorA
Creates a new Cognigy.AI knowledge connector for automated content ingestion from external sources like SharePoint or Confluence. MUTATING: Set dryRun=false to create.
Inputs: name (required), type (required), version (required), extension (required), +4 more.
create_knowledge_sourceA
Creates a new Cognigy.AI knowledge source for RAG content ingestion. Sources can be URLs, uploaded files, or manual text. Content is automatically chunked and embedded. MUTATING: Set dryRun=false to create.
Inputs: name (required), knowledgeStoreId (required), url, text, +4 more.
create_knowledge_storeA
Creates a new Cognigy.AI knowledge store for RAG content. Knowledge stores contain sources (documents) that AI Agents can search to answer questions. MUTATING: Set dryRun=false to create.
Inputs: name (required), projectId (required), dryRun, language, +4 more.
create_llmA
Creates a new Cognigy.AI large language model configuration. LLMs power Knowledge AI, AI Agents, and generative features. Requires a connection with provider credentials. MUTATING: Set dryRun=false to create.
Inputs: name (required), provider (required), modelType (required), projectId (required), +6 more.
create_nlu_connectorA
Creates a new Cognigy.AI NLU connector for external NLU service integration. Supports Dialogflow, LUIS, Watson, Alexa, Lex, and custom code connectors. MUTATING: Set dryRun=false to create.
Inputs: name (required), type (required), projectId (required), dryRun, +1 more.
create_nodeA
Creates a new node in a Cognigy.AI flow. MUTATING: This modifies the flow. Use dryRun=true (default) to validate first. Nodes are the building blocks of conversation logic (Say, Question, If, Code, etc.).
Inputs: type (required), flowId (required), targetNodeId (required), mode, +6 more.
create_packageA
Creates a package from selected resources in a Cognigy.AI project. Packages bundle flows, endpoints, and other resources for transfer between projects. MUTATING: Set dryRun=false to create. Async operation.
Inputs: name (required), projectId (required), resourceIds (required), dryRun, +3 more.
create_package_download_linkA
Creates a temporary download link for a Cognigy.AI package. The link can be used to download the package file for backup or transfer. Links are time-limited.
Inputs: packageId (required).
create_sentenceA
Creates a new example sentence for Cognigy.AI NLU intent training. MUTATING: This modifies the intent's training data. Use dryRun=true (default) to validate first. After creating, call train_intents to retrain.
Inputs: text (required), flowId (required), intentId (required), localeId (required), +1 more.
create_snapshotA
Creates a snapshot of a Cognigy.AI project. Snapshots capture the entire project configuration (flows, intents, endpoints, etc.) for backup or deployment. MUTATING: Set dryRun=false to create. Async operation - polls until complete.
Inputs: name (required), projectId (required), dryRun, timeoutMs, +2 more.
create_snapshot_download_linkA
Creates a temporary download link for a Cognigy.AI snapshot. The link can be used to download the snapshot as a file for backup or transfer to another environment. Links are time-limited.
Inputs: snapshotId (required).
delete_connectionADestructiveIdempotent
Deletes a Cognigy.AI connection. WARNING: This is destructive and cannot be undone. Flows using this connection will break. MUTATING: Set dryRun=false to delete.
Inputs: connectionId (required), dryRun.
delete_contact_profileADestructiveIdempotent
Deletes a Cognigy.AI contact profile. WARNING: This permanently removes all stored user data for this profile. MUTATING: Set dryRun=false to delete.
Inputs: profileId (required), dryRun.
delete_extensionADestructiveIdempotent
Deletes a Cognigy.AI Extension. WARNING: Flows using nodes from this extension will break. MUTATING: Set dryRun=false to delete.
Inputs: extensionId (required), dryRun.
delete_functionADestructiveIdempotent
Deletes a Cognigy.AI Function. WARNING: This is destructive. Flows calling this function will fail. MUTATING: Set dryRun=false to delete.
Inputs: functionId (required), dryRun.
delete_handover_providerADestructiveIdempotent
Deletes a Cognigy.AI handover provider. WARNING: Endpoints using this provider will no longer be able to escalate to live agents. MUTATING: Set dryRun=false to delete.
Inputs: projectId (required), providerId (required), dryRun.
delete_intentADestructiveIdempotent
Deletes an intent from a Cognigy.AI flow. MUTATING & DESTRUCTIVE: This permanently removes the intent and its example sentences. Use dryRun=true (default) to validate first. After deleting, call train_intents to retrain.
Inputs: flowId (required), intentId (required), dryRun.
delete_knowledge_chunkADestructiveIdempotent
Deletes a Cognigy.AI knowledge chunk. The content will no longer be searchable via RAG. MUTATING: Set dryRun=false to delete.
Inputs: chunkId (required), sourceId (required), knowledgeStoreId (required), dryRun.
delete_knowledge_connectorADestructiveIdempotent
Deletes a Cognigy.AI knowledge connector. Stops automated content ingestion from the external source. MUTATING: Set dryRun=false to delete.
Inputs: connectorId (required), knowledgeStoreId (required), dryRun.
delete_knowledge_sourceADestructiveIdempotent
Deletes a Cognigy.AI knowledge source and all its chunks. WARNING: This is destructive. The document content will no longer be searchable. MUTATING: Set dryRun=false to delete.
Inputs: sourceId (required), knowledgeStoreId (required), dryRun.
delete_knowledge_storeADestructiveIdempotent
Deletes a Cognigy.AI knowledge store and ALL its sources and chunks. WARNING: This is destructive and cannot be undone. AI Agents using this store will lose access. MUTATING: Set dryRun=false to delete.
Inputs: knowledgeStoreId (required), dryRun.
delete_llmADestructiveIdempotent
Deletes a Cognigy.AI large language model configuration. WARNING: Features using this LLM will stop working. MUTATING: Set dryRun=false to delete.
Inputs: largeLanguageModelId (required), dryRun.
delete_nlu_connectorADestructiveIdempotent
Deletes a Cognigy.AI NLU connector. WARNING: Endpoints using this connector will lose NLU functionality. MUTATING: Set dryRun=false to delete.
Inputs: nluConnectorId (required), dryRun.
delete_nodeADestructiveIdempotent
Deletes a node from a Cognigy.AI flow. MUTATING & DESTRUCTIVE: This permanently removes the node. Use dryRun=true (default) to validate first. Child nodes may also be affected.
Inputs: flowId (required), nodeId (required), dryRun.
delete_packageADestructiveIdempotent
Deletes a package from a Cognigy.AI project. DESTRUCTIVE & IRREVERSIBLE: The package will be permanently removed. Use dryRun=true (default) to validate first. Async operation.
Inputs: packageId (required), dryRun, timeoutMs, pollIntervalMs.
delete_snapshotADestructiveIdempotent
Deletes a snapshot from a Cognigy.AI project. DESTRUCTIVE & IRREVERSIBLE: The snapshot and all its data will be permanently removed. Use dryRun=true (default) to validate first. Async operation.
Inputs: snapshotId (required), dryRun, timeoutMs, pollIntervalMs.
diff_snapshotsARead-onlyIdempotent
Compares two Cognigy.AI snapshots and shows what changed (added, removed, modified resources). Useful for reviewing changes before deployment or understanding what a snapshot update will affect.
Inputs: snapshotIdA (required), snapshotIdB (required), resourceTypes.
export_contact_profileARead-onlyIdempotent
Exports all data for a Cognigy.AI contact profile. Use this for GDPR data access requests. Returns all stored profile data in a portable format.
Inputs: profileId (required).
generate_nlu_scoresARead-onlyIdempotent
Scores a test utterance against a Cognigy.AI flow's trained NLU intents. Returns ranked intent matches with confidence scores. Use this to test NLU recognition before deployment.
Inputs: sentence (required), projectId (required), flowReferenceId (required), localeReferenceId (required).
generate_node_outputARead-only
Uses Cognigy's generative AI to create content for Say nodes. Generates either plain text messages or rich Adaptive Cards based on a natural language prompt. Returns content you can use with create_node or update_node.
Inputs: flowId (required), prompt (required), localeId (required), lastOutput, +2 more.
generate_sentencesA
Uses Cognigy AI to generate example sentences for an intent. The generated sentences are NOT automatically added - use create_sentence to add them. Useful for quickly expanding NLU training data.
Inputs: flowId (required), intentId (required), limit, localeId.
get_audit_eventARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI audit event. Returns the full change details including before/after values.
Inputs: auditEventId (required).
get_call_metricsARead-onlyIdempotent
Gets Cognigy.AI call counter metrics (Voice Gateway) for one month. Returns aggregated call counts for a project or the entire organization.
Inputs: year, month, projectId.
get_connectionARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI connection. Returns connection metadata and schema. NOTE: Secret values are REDACTED for security - this tool only shows field names, not actual credentials.
Inputs: connectionId (required).
get_contact_profileARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI contact profile. Returns stored user data, goals, and profile metadata.
Inputs: profileId (required).
get_contact_profile_schemaARead-onlyIdempotent
Gets the contact profile schema for a Cognigy.AI project. The schema defines what custom fields can be stored in contact profiles.
Inputs: projectId (required).
get_conversationARead-onlyIdempotent
Gets conversation details for a specific Cognigy.AI session. Returns all inputs/outputs, timestamps, and metadata for the session. Use this to analyze a complete conversation thread.
Inputs: sessionId (required), projectId.
get_conversation_metricsARead-onlyIdempotent
Gets Cognigy.AI conversation counter metrics for one month. Returns aggregated conversation counts for a project or the entire organization.
Inputs: year, month, projectId.
get_conversationsARead-onlyIdempotent
Gets conversations for specific contacts in a Cognigy.AI project. Returns conversation history including inputs, outputs, and metadata. Use this to analyze user interactions.
Inputs: projectId (required), contactIds (required).
get_endpointARead-onlyIdempotent
Gets detailed configuration of a specific Cognigy.AI endpoint. Returns channel settings, flow/agent binding, and runtime configuration. Use this to inspect endpoint behavior.
Inputs: endpointId (required).
get_extensionARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI Extension. Returns package info, available nodes, connections schemas, and settings.
Inputs: extensionId (required).
get_flowARead-onlyIdempotent
Gets detailed metadata about a specific Cognigy.AI flow. Returns flow configuration, locale info, and timestamps. Use this to inspect a flow before modifying it.
Inputs: flowId (required).
get_flow_settingsARead-onlyIdempotent
Gets the settings/configuration of a Cognigy.AI flow. Returns NLU settings, thresholds, and other flow-level configurations. Use this before updating flow settings.
Inputs: flowId (required).
get_functionARead-onlyIdempotent
Gets detailed configuration of a specific Cognigy.AI Function. Returns the function code, settings, and runtime configuration.
Inputs: functionId (required).
get_function_instanceARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI Function instance. Returns execution status, timing, input parameters, and output/error.
Inputs: functionId (required), functionInstanceId (required).
get_handover_providerARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI handover provider. Returns provider type, configuration, and connection settings.
Inputs: providerId (required).
get_handover_serviceARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI handover service (accepts the service ID or name). Returns its version, service URL, and properties.
Inputs: serviceId (required).
get_intentARead-onlyIdempotent
Gets detailed configuration of a specific intent in a Cognigy.AI flow. Returns the intent's conditions, rules, confirmation sentences, and settings. Use this to inspect NLU behavior before modifying.
Inputs: flowId (required), intentId (required), localeId.
get_knowledge_chunkARead-onlyIdempotent
Gets the full content of a specific Cognigy.AI knowledge chunk. Returns the complete text, metadata, and source information. Use this to inspect what content is being used in RAG searches.
Inputs: chunkId (required), sourceId (required), knowledgeStoreId (required).
get_knowledge_connectorARead-onlyIdempotent
Gets detailed configuration of a specific Cognigy.AI knowledge connector. Returns connector type, schedule, connection settings, and run status.
Inputs: connectorId (required), knowledgeStoreId (required).
get_knowledge_query_metricsARead-onlyIdempotent
Gets Cognigy.AI Knowledge AI query counter metrics for one month. Returns aggregated knowledge query counts for a project or the entire organization.
Inputs: year, month, projectId.
get_knowledge_sourceARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI knowledge source. Returns source metadata, processing status, chunk count, and ingestion details.
Inputs: sourceId (required), knowledgeStoreId (required).
get_knowledge_storeARead-onlyIdempotent
Gets detailed configuration of a specific Cognigy.AI knowledge store. Returns store settings, language, embedding model, and source counts.
Inputs: knowledgeStoreId (required).
get_latest_log_entriesARead-onlyIdempotent
Gets the latest execution log entries from a Cognigy.AI project. Use this for debugging flow execution, viewing errors, or monitoring agent behavior.
Inputs: projectId (required), type, limit, flowName.
get_llmARead-onlyIdempotent
Gets detailed configuration of a specific Cognigy.AI large language model. Returns provider settings, model type, connection details, and fallback configuration.
Inputs: largeLanguageModelId (required).
get_nlu_connectorARead-onlyIdempotent
Gets detailed configuration of a specific Cognigy.AI NLU connector. Returns type, settings, and connection details for external NLU service integration.
Inputs: nluConnectorId (required).
get_nodeARead-onlyIdempotent
Gets detailed configuration of a specific node in a Cognigy.AI flow. Returns the node's type, label, config fields, and settings. Use this to inspect node behavior before modifying it.
Inputs: flowId (required), nodeId (required), localeId.
get_node_descriptorsARead-onlyIdempotent
Gets all available node types (blueprints) that can be created in a Cognigy.AI flow. Returns node type definitions including their fields, appearance, and constraints. Use this to understand what nodes can be added to a flow.
Inputs: flowId (required).
get_nodesARead-onlyIdempotent
Lists all nodes in a Cognigy.AI flow. Nodes are the building blocks of conversation logic (Say, Question, If, Code, etc.). Use this to explore flow structure before reading specific nodes or modifying the flow.
Inputs: flowId (required), limit.
get_packageARead-onlyIdempotent
Gets detailed information about a Cognigy.AI package including its name, description, and contained resources.
Inputs: packageId (required).
get_playbookARead-onlyIdempotent
Gets detailed Cognigy.AI playbook configuration including all steps and assertions. Use this to understand what a playbook tests before running it.
Inputs: playbookId (required).
get_playbook_runARead-onlyIdempotent
Gets detailed results of a Cognigy.AI playbook run including step-by-step assertion outcomes. Use this to analyze test failures and debug conversation flows.
Inputs: playbookId (required), playbookRunId (required).
get_snapshotARead-onlyIdempotent
Gets detailed information about a specific Cognigy.AI snapshot. Returns name, description, hash, and packaging status. Use this to inspect a version before deployment.
Inputs: snapshotId (required).
get_snapshot_resourcesARead-onlyIdempotent
Lists resources (flows, locales, NLU connectors, LLMs) contained in a Cognigy.AI snapshot. Use this to inspect what a snapshot contains before restoring or to compare versions.
Inputs: snapshotId (required), resourceType (required), limit.
get_taskARead-onlyIdempotent
Gets detailed status of a specific Cognigy.AI async task. Returns progress, status, and failure reason if applicable. Use this to poll long-running operations to completion.
Inputs: taskId (required), projectId.
get_transcriptARead-onlyIdempotent
Assembles a human-readable transcript for a Cognigy.AI session. Shows the conversation flow between user and bot in chronological order. Use this for reviewing conversation quality or debugging.
Inputs: sessionId (required), format, projectId.
inject_contextA
Injects context data into a Cognigy.AI session. Context is shared state accessible by flow nodes. Use this to set user data, preferences, or state before/during conversations.
Inputs: userId (required), context (required), sessionId (required).
list_audit_eventsARead-onlyIdempotent
Lists Cognigy.AI audit events. Audit events track all changes made to resources (flows, intents, endpoints, etc.) by users. Useful for compliance and debugging.
Inputs: skip, user, limit, eventType, +1 more.
list_connectionsARead-onlyIdempotent
Lists Cognigy.AI connections (external service integrations like databases, APIs, etc.). Connections store credentials securely. Use this to find available connections for a project or organization.
Inputs: skip, limit, filter, projectId, +1 more.
list_contact_profilesBRead-onlyIdempotent
Lists Cognigy.AI contact profiles. Contact profiles store user data across sessions (name, preferences, conversation history metadata).
Inputs: skip, limit, filter, projectId.
list_endpointsARead-onlyIdempotent
Lists all endpoints in a Cognigy.AI project. Endpoints are channel connectors (Webchat, REST, Voice, etc.) that expose flows/agents to users. Use this to discover deployed channels.
Inputs: projectId (required), limit.
list_extensionsARead-onlyIdempotent
Lists Cognigy.AI Extensions. Extensions are custom node packages that add new capabilities to flows (e.g., integrations, custom actions).
Inputs: skip, limit, filter, projectId.
list_flowsARead-onlyIdempotent
Lists all flows in a Cognigy.AI project. Flows are conversation logic containers. Use this to discover flows before reading or modifying them.
Inputs: projectId (required), limit.
list_function_instancesARead-onlyIdempotent
Lists running and completed instances of a Cognigy.AI Function. Shows execution history, status, and results.
Inputs: functionId (required), skip, limit.
list_functionsBRead-onlyIdempotent
Lists Cognigy.AI Functions. Functions are custom code modules that can be triggered to run computations, integrations, or scheduled jobs outside of flow execution.
Inputs: skip, limit, filter, projectId.
list_handover_providersARead-onlyIdempotent
Lists Cognigy.AI handover providers. Handover providers enable live agent escalation (e.g., Salesforce, Genesys, RingCentral).
Inputs: skip, limit, projectId.
list_handover_servicesBRead-onlyIdempotent
Lists available Cognigy.AI handover services. Handover services are the supported integrations for live agent escalation.
Inputs: skip, limit.
list_intentsARead-onlyIdempotent
Lists all intents in a Cognigy.AI flow. Intents are the NLU triggers that match user utterances to flow logic. Use this to explore NLU configuration before training or modifying intents.
Inputs: flowId (required), limit, localeId, includeChildren.
list_knowledge_chunksARead-onlyIdempotent
Lists knowledge chunks in a Cognigy.AI knowledge store. Chunks are the actual text segments used for RAG retrieval, created by splitting source documents.
Inputs: knowledgeStoreId (required), skip, limit, filter, +1 more.
list_knowledge_connectorsARead-onlyIdempotent
Lists Cognigy.AI knowledge connectors for automated content ingestion. Connectors can pull content from external sources like SharePoint, Confluence, or custom APIs.
Inputs: knowledgeStoreId (required), skip, limit, filter.
list_knowledge_sourcesARead-onlyIdempotent
Lists knowledge sources in a Cognigy.AI knowledge store. Sources are documents (PDFs, web pages, text files) that have been ingested and chunked for RAG retrieval.
Inputs: knowledgeStoreId (required), skip, limit, filter.
list_knowledge_storesARead-onlyIdempotent
Lists Cognigy.AI Knowledge AI stores. Knowledge stores are containers for RAG (Retrieval-Augmented Generation) content used by AI Agents to answer questions from your data.
Inputs: skip, limit, filter, projectId.
list_llmsARead-onlyIdempotent
Lists Cognigy.AI large language model configurations. LLMs are used for generative AI features like Knowledge AI, AI Agents, and node output generation. Shows provider, model type, and connection info.
Inputs: skip, limit, filter, projectId, +1 more.
list_nlu_connectorsARead-onlyIdempotent
Lists Cognigy.AI NLU connectors. NLU connectors enable integration with external NLU services like Dialogflow, LUIS, Watson, or custom solutions for intent recognition.
Inputs: skip, limit, filter, projectId.
list_packagesARead-onlyIdempotent
Lists packages in a Cognigy.AI project. Packages are portable bundles of resources (flows, intents, etc.) that can be transferred between projects or environments.
Inputs: limit, projectId.
list_playbook_runsARead-onlyIdempotent
Lists Cognigy.AI playbook run history showing pass/fail status, timestamps, and run metadata. Use this to review test results over time.
Inputs: playbookId (required), limit.
list_playbooksARead-onlyIdempotent
Lists all playbooks in a Cognigy.AI project. Playbooks are automated test scenarios with steps and assertions for testing conversational flows.
Inputs: projectId (required), limit.
list_projectsARead-onlyIdempotent
Lists all Cognigy.AI projects accessible by your API key. Use this to discover available projects before working with flows, intents, or other resources.
Inputs: limit.
list_sentencesARead-onlyIdempotent
Lists example sentences (training data) for a Cognigy.AI NLU intent. Use this to review training data quality before training.
Inputs: flowId (required), intentId (required), limit.
list_snapshotsARead-onlyIdempotent
Lists all snapshots in a Cognigy.AI project. Snapshots are versioned backups of project configuration used for deployment and rollback. Use this to see available versions.
Inputs: projectId (required), limit.
list_tasksARead-onlyIdempotent
Lists async tasks in Cognigy.AI. Tasks track long-running operations like snapshot creation, training, and imports. Use this to monitor background job status.
Inputs: limit, projectId.
merge_contact_profilesA
Merges the Cognigy.AI contact profile of a given contact ID into a target profile, combining their data and contact IDs. Use when the same user has multiple profiles. MUTATING: Set dryRun=false to merge.
Inputs: projectId (required), sourceContactId (required), targetProfileId (required), dryRun.
merge_packageA
Merges a package into a Cognigy.AI project, importing selected resources. Use localeMapping to map package locales to project locales. KNOWN LIMITATION: in live testing (Cognigy 2026.20) every merge failed with 'Resource with id ... is not in the source', and the public API offers no way to list a package's resources; if this happens, import the package in the Cognigy.AI UI instead. MUTATING: Set dryRun=false to merge. Async operation.
Inputs: packageId (required), resourceIds (required), localeMapping (required), dryRun, +3 more.
move_nodeA
Moves a node to a new position in a Cognigy.AI flow. MUTATING: This reorganizes the flow structure. Use dryRun=true (default) to validate first. Moving nodes affects execution order.
Inputs: mode (required), flowId (required), nodeId (required), targetNodeId (required), +2 more.
package_snapshotA
Packages a Cognigy.AI snapshot for download or transfer. Creates a downloadable package from the snapshot. Use create_snapshot_download_link after packaging to get the download URL. MUTATING: Set dryRun=false to package. Async operation.
Inputs: snapshotId (required), dryRun, timeoutMs, pollIntervalMs.
promote_snapshotA
Promotes a Cognigy.AI snapshot for deployment by packaging it and generating a download link. Use this to prepare a snapshot for transfer to another environment. MUTATING: Set dryRun=false to package. Async operation.
Inputs: snapshotId (required), dryRun, timeoutMs, pollIntervalMs.
remove_contact_idADestructiveIdempotent
Removes a contact ID from a Cognigy.AI contact profile. Use this when a user identifier should no longer be associated with a profile. MUTATING: Set dryRun=false to remove.
Inputs: contactId (required), profileId (required), dryRun.
reset_contextAIdempotent
Resets the context for a Cognigy.AI session, clearing all stored state. Use this to start a fresh conversation or clear user data during testing.
Inputs: userId (required), sessionId (required), entrypoint (required), flowReferenceId (required).
restore_snapshotADestructive
Restores a snapshot to its Cognigy.AI project, replacing the current configuration. DESTRUCTIVE: Current project state will be overwritten with the snapshot's state. Use dryRun=true (default) to validate first. Async operation.
Inputs: snapshotId (required), dryRun, timeoutMs, pollIntervalMs.
run_knowledge_connectorA
Triggers a Cognigy.AI knowledge connector to run immediately. Pulls content from the external source and creates/updates knowledge chunks. MUTATING: Set dryRun=false to run.
Inputs: connectorId (required), knowledgeStoreId (required), dryRun.
run_playbookA
Runs a Cognigy.AI playbook test scenario against a flow. MUTATING: This executes test assertions. Use dryRun=true (default) to validate first. Returns pass/fail results with assertion details.
Inputs: flowId (required), localeId (required), entrypoint (required), playbookId (required), +3 more.
run_regressionA
Runs all Cognigy.AI playbooks in a project as a regression test suite. MUTATING: This executes tests. Use dryRun=true (default) to preview. Returns pass/fail summary with failing playbooks highlighted.
Inputs: flowId (required), localeId (required), projectId (required), dryRun, +2 more.
score_utteranceARead-onlyIdempotent
Scores a test utterance against a Cognigy.AI flow's trained NLU intents. Returns the best matching intent with confidence score. Use this to quickly test if an utterance would be recognized correctly.
Inputs: flowId (required), projectId (required), utterance (required), localeId, +1 more.
search_nodesARead-onlyIdempotent
Searches for nodes in a Cognigy.AI flow by text content. Finds nodes containing the search term in their configuration (messages, conditions, code, etc.). Use this to locate specific content within large flows.
Inputs: query (required), flowId (required), localeId (required).
search_resourcesARead-onlyIdempotent
Searches across all Cognigy.AI resources in the organization. Finds flows, intents, endpoints, functions, playbooks, and more by name or content. Powerful for discovering resources across projects.
Inputs: query (required), skip, limit, projectId, +1 more.
set_contact_profile_schemaAIdempotent
Adds or updates custom fields in the Cognigy.AI contact profile schema of a project. Built-in fields (firstname, email, ...) always exist; existing custom fields not listed are kept. MUTATING: Set dryRun=false to update.
Inputs: schema (required), projectId (required), dryRun.
stop_function_instanceA
Stops a running Cognigy.AI Function instance. Use this to cancel a long-running or stuck function. MUTATING: Set dryRun=false to stop.
Inputs: functionId (required), functionInstanceId (required), dryRun.
test_llm_connectionA
Tests the connection to a Cognigy.AI large language model provider. Validates that the credentials are correct and the provider is reachable. Use this to verify LLM setup before using it in flows.
Inputs: largeLanguageModelId (required).
train_intentsA
Trains the NLU model for a Cognigy.AI flow. MUTATING: This triggers model training. Use dryRun=true (default) to validate first. Training is async - this tool polls until completion or timeout.
Inputs: flowId (required), mode, dryRun, localeId, +2 more.
trigger_functionA
Triggers a Cognigy.AI Function to run immediately. Creates a new function instance that executes the function code. MUTATING: Set dryRun=false to trigger.
Inputs: functionId (required), input, dryRun.
unmerge_contact_profilesA
Splits a merged Cognigy.AI contact profile back into separate profiles. Use when profiles were incorrectly merged. MUTATING: Set dryRun=false to unmerge.
Inputs: contactId (required), profileId (required), dryRun.
update_analytics_recordAIdempotent
Updates Cognigy.AI analytics records for a session. Use this to add custom tracking properties to conversation analytics. MUTATING: Set dryRun=false to update.
Inputs: contactId (required), projectId (required), sessionId (required), properties (required), +2 more.
update_connectionAIdempotent
Updates an existing Cognigy.AI connection. Use this to change connection name or update credential values. MUTATING: Set dryRun=false to update.
Inputs: connectionId (required), name, dryRun, fields.
update_contact_profileAIdempotent
Updates an existing Cognigy.AI contact profile. Use this to modify stored user data or GDPR consent. MUTATING: Set dryRun=false to update.
Inputs: profileId (required), dryRun, profile, acceptedGDPR.
update_extensionAIdempotent
Updates Cognigy.AI Extension settings like trusted code flag. Use this to enable/disable full Node.js API access. MUTATING: Set dryRun=false to update.
Inputs: extensionId (required), dryRun, trustedCode.
update_extension_packageAIdempotent
Updates a Cognigy.AI Extension with a new package version from a URL. Use this to upgrade an extension to a new version. MUTATING: Set dryRun=false to update.
Inputs: url (required), extensionId (required), dryRun.
update_functionAIdempotent
Updates an existing Cognigy.AI Function. Use this to change its name or code, or to enable/disable it. MUTATING: Set dryRun=false to update.
Inputs: functionId (required), code, name, dryRun, +1 more.
update_handover_providerAIdempotent
Updates an existing Cognigy.AI handover provider. Use this to rename it or change its properties. MUTATING: Set dryRun=false to update.
Inputs: projectId (required), providerId (required), name, dryRun, +1 more.
update_intentAIdempotent
Updates an existing intent in a Cognigy.AI flow. MUTATING: This modifies the intent. Use dryRun=true (default) to validate first. After updating, call train_intents to retrain the NLU model.
Inputs: flowId (required), intentId (required), name, tags, +8 more.
update_knowledge_chunkAIdempotent
Updates an existing Cognigy.AI knowledge chunk. If text is changed, the chunk will be re-embedded. MUTATING: Set dryRun=false to update.
Inputs: chunkId (required), sourceId (required), knowledgeStoreId (required), text, +1 more.
update_knowledge_connectorAIdempotent
Updates an existing Cognigy.AI knowledge connector. Use this to change settings or name. MUTATING: Set dryRun=false to update.
Inputs: connectorId (required), knowledgeStoreId (required), name, dryRun.
update_knowledge_sourceAIdempotent
Updates an existing Cognigy.AI knowledge source. Use this to change name or description. MUTATING: Set dryRun=false to update.
Inputs: sourceId (required), knowledgeStoreId (required), name, dryRun, +1 more.
update_knowledge_storeAIdempotent
Updates an existing Cognigy.AI knowledge store. Use this to change name or description. MUTATING: Set dryRun=false to update.
Inputs: knowledgeStoreId (required), name, dryRun, description.
update_llmAIdempotent
Updates an existing Cognigy.AI large language model configuration. Use this to change name, description, credentials, or provider settings. MUTATING: Set dryRun=false to update.
Inputs: largeLanguageModelId (required), name, dryRun, isDefault, +3 more.
update_nlu_connectorAIdempotent
Updates an existing Cognigy.AI NLU connector. Use this to change name or update type-specific settings. MUTATING: Set dryRun=false to update.
Inputs: nluConnectorId (required), name, dryRun, settings.
update_nodeAIdempotent
Updates an existing node in a Cognigy.AI flow. MUTATING: This modifies the node. Use dryRun=true (default) to validate first. Only provided fields are updated; others remain unchanged.
Inputs: flowId (required), nodeId (required), label, config, +5 more.
upload_extensionA
Uploads a new Cognigy.AI Extension from a URL. The extension package must be a .tar.gz file. This is an async operation that polls until complete. MUTATING: Set dryRun=false to upload.
Inputs: url (required), projectId (required), dryRun.
upload_packageA
Uploads a package file to a Cognigy.AI project. Use this to import a previously downloaded package or transfer resources between environments. MUTATING: Set dryRun=false to upload. Async operation.
Inputs: filePath (required), projectId (required), dryRun, timeoutMs, +1 more.
upload_snapshot_packageA
Uploads a snapshot package file to a Cognigy.AI project. Use this to restore a previously downloaded snapshot or transfer a snapshot between environments. MUTATING: Set dryRun=false to upload. Async operation.
Inputs: filePath (required), projectId (required), dryRun, timeoutMs, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 138 tools
Most tools have distinct purposes, but several pairs overlap noticeably (e.g., score_utterance vs generate_nlu_scores both score utterances against NLU intents; create_package, package_snapshot, and promote_snapshot all package resources). Descriptions help, but with 138 tools the agent can still misselect among similar-sounding operations.
Every tool follows a consistent snake_case verb_noun pattern (e.g., list_flows, get_node, create_intent, delete_knowledge_store). Minor variations like get_nodes vs get_node are natural, and no mixed conventions appear.
138 tools far exceeds reasonable bounds for an agent-facing server, even given Cognigy.AI's broad platform. This extreme count creates a high cognitive load and increases selection errors, matching the rubric's 50+ tools criterion for an extreme mismatch.
Coverage is broad across many resources (NLU, knowledge, contacts, snapshots, packages, functions, LLMs), but notable gaps exist: no create_flow/update_flow/delete_flow, no create_endpoint/update_endpoint/delete_endpoint, and no update_flow_settings. These missing core CRUD operations for central resources limit end-to-end management workflows.
Maintenance
Related MCP Connectors
Marketo MCP server for AI. 130 tools to operate Marketo from Claude, Cursor, or ChatGPT.
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
Email inboxes, webhook URLs, mock APIs and 170+ tools for coding agents over one MCP connection.
Related MCP Servers
- FlicenseBqualityDmaintenanceA Model Context Protocol server that enables Claude users to access specialized OpenAI agents (web search, file search, computer actions) and a multi-agent orchestrator through the MCP protocol.410-
- AlicenseDqualityDmaintenanceA foundation for building custom local Model Context Protocol (MCP) servers that provide tools accessible to AI assistants like Cursor or Claude Desktop.137MIT
- FlicenseNot gradedqualityDmaintenanceA Model Context Protocol (MCP) server built with the mcp-framework for developing and managing custom tools. It provides a structured foundation for building and integrating modular components like data processors and API clients into Claude Desktop.-
- AlicenseAqualityDmaintenanceA Model Context Protocol server providing tools for DB queries, API calls, file I/O, and text transformations, enabling AI agents like Claude to perform real-world actions.10MIT