Kimi AI n8n Node

This is a custom n8n node for interacting with the Kimi AI (Moonshot AI) API, designed to provide a seamless and powerful integration for conversational AI workflows. It supports chat completions, multi-turn conversations with robust memory management, streaming responses, and advanced AI features.

Features

• Chat Completions: Engage in dynamic conversations with the Kimi AI model.
• Multi-turn Conversations: Maintain context across multiple interactions with internal memory management.
• Persistent Memory Handling: Supports session-based memory to keep track of conversation history.
• Streaming Responses: Receive responses in real-time for a more interactive user experience.
• System Prompts: Guide the AI's behavior with customizable system prompts.
• JSON Mode: Obtain structured JSON responses for easier parsing and integration.
• Tool/Function Calling Preparation: Designed with future tool/function calling capabilities in mind.
• Long Context Optimization: Handles longer conversations efficiently.
• Workflow-friendly AI Automation: Built for integration into various n8n workflows, including chatbots and autonomous agents.
• Model Listing: Easily retrieve a list of available Kimi AI models.

Installation

To install this custom node, you will need to have n8n installed and running. Follow these steps:

1. Clone the repository (or download the source code):

   git clone https://github.com/manus-ai/n8n-nodes-kimi-ai.git
   cd n8n-nodes-kimi-ai

2. Install dependencies:

   npm install

3. Build the node:

   npm run build

4. Link the node to your n8n instance:

   npm link
   # Then, in your n8n installation directory:
   npm link n8n-nodes-kimi-ai

5. Restart n8n to load the new node.

Credentials

This node uses API Key authentication.

1. In n8n, go to Credentials.
2. Add a new credential of type Kimi AI API.
3. Enter your API Key obtained from the Kimi Open Platform.
   • The API Key will be used in the Authorization: Bearer {{API_KEY}} header.

Usage

Chat Completion

1. Drag and drop the Kimi AI node onto your workflow.
2. Select Resource: Chat.
3. Select Operation: Chat Completion.
4. Provide a User Message.
5. (Optional) Configure System Prompt to guide the AI.
6. (Optional) Choose a Memory Mode:
   • No Memory: Each request is independent.
   • Session Memory: Maintains conversation history based on a Session ID.
7. (Optional) Adjust Options such as Temperature, Max Tokens, JSON Mode, and Stream.

Model Listing

1. Drag and drop the Kimi AI node onto your workflow.
2. Select Resource: Model.
3. Select Operation: List.
4. Execute the node to get a list of available Kimi AI models.

Memory Management

The Kimi API is stateless. This node implements internal conversation memory management to support multi-turn conversations. When Session Memory is enabled, the node will:

• Automatically manage the messages array, including user and assistant history.
• Implement a rolling context window to keep the conversation within the model's token limits.
• Ensure system prompts persist across turns.

Streaming Support

When the Stream option is enabled for chat completions, the node will:

• Handle SSE (Server-Sent Events) streaming.
• Process partial token streaming and aggregate chunks for progressive output.

Error Handling

The node includes robust error handling for common API issues such as:

• Invalid API keys
• Rate limits
• Token overflow
• Malformed messages
• Streaming interruptions
• Timeout errors

User-friendly n8n errors will be returned for easier debugging.

Advanced AI Features

• JSON Mode: Request responses in a structured JSON format.
• AI Agent Compatibility: Designed to work seamlessly within AI agent workflows.
• Function/Tool Calling Preparation: Architectural support for future integration of tool calling.
• Context Optimization: Strategies to manage and optimize conversation context.

Deliverables

This project provides the following files:

• packages/nodes/KimiAI/KimiAI.node.ts: The main n8n node definition.
• packages/credentials/KimiApiCredentials.ts: API key credential definition.
• packages/memoryManager/MemoryManager.ts: Logic for managing conversation history.
• packages/streamingParser/StreamingParser.ts: Handles parsing of streaming responses.
• packages/utils/Transport.ts: Reusable transport layer for API requests.
• packages/resources/ChatResource.ts: Defines chat-related operations and fields.
• packages/resources/ModelResource.ts: Defines model-related operations and fields.
• packages/nodes/KimiAI/kimi.svg: Node icon.
• package.json: Project dependencies and scripts.
• tsconfig.json: TypeScript configuration.
• README.md: This documentation file.

Future Enhancements

Potential future features include:

• Conversation export/import.
• Automatic summarization of old context.
• Embedding-ready architecture and Vector DB compatibility.
• Full AI Agent mode and Tool calling abstraction layer.
• Prompt templates.
• Binary and file support (PDF context, text files, RAG ingestion, file summarization).

Author: Manus AI Homepage: https://www.manus.im Repository: https://github.com/manus-ai/n8n-nodes-kimi-ai