ds-mcp-flowise 1.1.3 → 1.1.5

package/README.md

@@ -1,256 +1,270 @@
-# DS-MCP-FLOWISE
-
-MCP (Model Context Protocol) server that gives AI assistants (Claude, GPT, etc.) deep knowledge of [Flowise](https://flowiseai.com/) nodes so they can help you design, build, and deploy chatflows and agentflows.
-
-## What This Does
-
-When connected to Claude Code (or other MCP clients), this server lets the AI:
-- Know all 311 Flowise nodes and their configurations
-- Understand how nodes connect together
-- Design complete flows based on your requirements
-- Output valid JSON you can import into Flowise
-- **Deploy flows directly to your Flowise instance** (optional API integration)
-
-## Quick Start
-
-### 1. Add to Claude Code
-
-Add this to your Claude Code MCP configuration (`claude_desktop_config.json` or via settings):
-
-**Windows:**
-```json
-{
-  "mcpServers": {
-    "flowise": {
-      "command": "cmd",
-      "args": ["/c", "npx", "-y", "ds-mcp-flowise"]
-    }
-  }
-}
-```
-
-**Mac / Linux:**
-```json
-{
-  "mcpServers": {
-    "flowise": {
-      "command": "npx",
-      "args": ["-y", "ds-mcp-flowise"]
-    }
-  }
-}
-```
-
-That's it. No cloning, no building. The `npx` command downloads and runs it automatically.
-
-### 2. Start a Conversation
-
-Open Claude Code and describe what you want to build:
-
-> "Build me a RAG chatbot that uses Pinecone for vector storage and OpenAI for the LLM"
-
-### 3. Get Your Flow
-
-Claude will use the MCP tools to:
-- Find the right nodes (ChatOpenAI, Pinecone, embeddings, etc.)
-- Check their schemas and required inputs
-- Design the complete flow
-- Output the JSON with all nodes and edges
-
-### 4. Import into Flowise
-
-1. Open Flowise
-2. Create a new Chatflow or Agentflow
-3. Click the menu (⋮) → **Load Chatflow**
-4. Paste the JSON Claude gave you
-5. Configure credentials (API keys) for each node
-6. Save and test
-
-## Flowise API Integration (Optional)
-
-Instead of manually importing JSON, you can connect directly to your Flowise instance to create, update, and manage chatflows via the API.
-
-### Setup
-
-1. Create a `.env` file in the project root (or in your working directory):
-
-```bash
-FLOWISE_API_URL=https://your-flowise-instance.com
-FLOWISE_API_KEY=your-api-key-here
-```
-
-2. Get your API key from Flowise:
-   - Open your Flowise instance
-   - Go to **Settings** → **API Keys**
-   - Create a new API key or copy an existing one
-
-3. The MCP server will automatically load credentials from `.env` when it starts.
-
-### Usage
-
-Once configured, you can ask Claude:
-
-- "Test the connection to my Flowise instance"
-- "List all my chatflows"
-- "Create a new chatflow with this design and deploy it"
-- "Update the existing chatflow with these changes"
-
-> **Note:** The Flowise API currently only supports **chatflows**. Agentflows must be created through the Flowise UI by importing the generated JSON.
-
-### Security Notes
-
-- Never commit your `.env` file to version control
-- The `.env` file is already in `.gitignore`
-- API keys should be kept private and rotated periodically
-
-## Alternative Installation Methods
-
-### Global Install (npm)
-
-```bash
-npm install -g ds-mcp-flowise
-```
-
-Then in your MCP config:
-
-**Windows:**
-```json
-{
-  "mcpServers": {
-    "flowise": {
-      "command": "cmd",
-      "args": ["/c", "ds-mcp-flowise"]
-    }
-  }
-}
-```
-
-**Mac / Linux:**
-```json
-{
-  "mcpServers": {
-    "flowise": {
-      "command": "ds-mcp-flowise"
-    }
-  }
-}
-```
-
-### Build from Source
-
-```bash
-git clone https://github.com/dtsoden/DS-MCP-FLOWISE.git
-cd DS-MCP-FLOWISE
-npm install
-npm run build
-```
-
-Then in your MCP config:
-
-**Windows:**
-```json
-{
-  "mcpServers": {
-    "flowise": {
-      "command": "node",
-      "args": ["C:\\path\\to\\DS-MCP-FLOWISE\\dist\\index.js"]
-    }
-  }
-}
-```
-
-**Mac / Linux:**
-```json
-{
-  "mcpServers": {
-    "flowise": {
-      "command": "node",
-      "args": ["/path/to/DS-MCP-FLOWISE/dist/index.js"]
-    }
-  }
-}
-```
-
-## Available Tools
-
-### Node Discovery
-
-| Tool | Description |
-|------|-------------|
-| `list_categories` | List all 26 node categories with counts |
-| `list_nodes` | List nodes, optionally filtered by category |
-| `get_node_schema` | Get detailed schema for a node including all inputs |
-| `search_nodes` | Search nodes by keyword |
-| `find_compatible_nodes` | Find nodes that can connect to a given node |
-
-### Template Library
-
-| Tool | Description |
-|------|-------------|
-| `list_templates` | List 64 marketplace templates by type |
-| `get_template` | Get complete template with nodes and edges |
-
-### Flow Building
-
-| Tool | Description |
-|------|-------------|
-| `validate_flow` | Validate a flow's nodes and connections |
-| `generate_flow_skeleton` | Generate a starting flow for common use cases |
-
-### Flowise API (requires .env configuration)
-
-| Tool | Description |
-|------|-------------|
-| `flowise_test_connection` | Test connection to your Flowise instance |
-| `flowise_list_chatflows` | List all chatflows in your instance |
-| `flowise_get_chatflow` | Get details of a specific chatflow |
-| `flowise_create_chatflow` | Create and deploy a new chatflow |
-| `flowise_update_chatflow` | Update an existing chatflow |
-| `flowise_delete_chatflow` | Delete a chatflow |
-
-> **Note:** The Flowise API currently only supports **chatflows**. Agentflows must be created through the Flowise UI by importing the generated JSON.
-
-## Example Prompts
-
-Once connected, try asking Claude:
-
-- "What vector stores does Flowise support?"
-- "Show me how to build a simple chatbot with memory"
-- "Create a RAG flow using Pinecone and Claude"
-- "What nodes can connect to a ConversationChain?"
-- "Build me an agent that can search the web and query a database"
-
-## What's Included
-
-- **311 Flowise nodes** with full schemas
-- **1,915 input parameters** documented
-- **26 node categories** (Chat Models, Vector Stores, Tools, etc.)
-- **64 marketplace templates** as examples
-- **SQLite database** for fast queries
-
-## Updating the Node Database
-
-To update with the latest Flowise nodes:
-
-```bash
-git clone https://github.com/dtsoden/DS-MCP-FLOWISE.git
-cd DS-MCP-FLOWISE
-npm install
-
-# This clones Flowise and extracts all node definitions
-npm run extract
-
-# This rebuilds the SQLite database
-npm run prepare-db
-
-npm run build
-```
-
-## Credits
-
-- [Flowise](https://github.com/FlowiseAI/Flowise) - The no-code LLM orchestration platform
-- Inspired by [n8n-mcp](https://github.com/czlonkowski/n8n-mcp) architecture
-
-## License
-
-MIT
+# DS-MCP-FLOWISE
+
+MCP (Model Context Protocol) server that gives AI assistants (Claude, GPT, etc.) deep knowledge of [Flowise](https://flowiseai.com/) nodes so they can help you design, build, and deploy chatflows and agentflows.
+
+## What This Does
+
+When connected to Claude Code (or other MCP clients), this server lets the AI:
+- Know all 311 Flowise nodes and their configurations
+- Understand how nodes connect together
+- Design complete flows based on your requirements
+- Output valid JSON you can import into Flowise
+- **Deploy flows directly to your Flowise instance** (optional API integration)
+
+## Quick Start
+
+### 1. Add to Claude Code
+
+Add this to your Claude Code MCP configuration (`claude_desktop_config.json` or via settings):
+
+**Windows:**
+```json
+{
+  "mcpServers": {
+    "flowise": {
+      "command": "cmd",
+      "args": ["/c", "npx", "-y", "ds-mcp-flowise"]
+    }
+  }
+}
+```
+
+**Mac / Linux:**
+```json
+{
+  "mcpServers": {
+    "flowise": {
+      "command": "npx",
+      "args": ["-y", "ds-mcp-flowise"]
+    }
+  }
+}
+```
+
+That's it. No cloning, no building. The `npx` command downloads and runs it automatically.
+
+### 2. Start a Conversation
+
+Open Claude Code and **explicitly tell it to use the flowise tools**:
+
+> "**Use the flowise tools** to build me a RAG chatbot that uses Pinecone for vector storage and OpenAI for the LLM. Give me the JSON to import into Flowise."
+
+**Important:** Without mentioning "flowise tools" or "flowise MCP", Claude may try to build a chat app from scratch instead of generating a Flowise chatflow.
+
+### 3. Get Your Flow
+
+Claude will use the MCP tools to:
+- Find the right nodes (ChatOpenAI, Pinecone, embeddings, etc.)
+- Check their schemas and required inputs
+- Design the complete flow
+- Output the JSON with all nodes and edges
+
+### 4. Import into Flowise
+
+1. Open Flowise
+2. Create a new Chatflow or Agentflow
+3. Click the menu (⋮) → **Load Chatflow**
+4. Paste the JSON Claude gave you
+5. Configure credentials (API keys) for each node
+6. Save and test
+
+## Flowise API Integration (Optional)
+
+Instead of manually importing JSON, you can connect directly to your Flowise instance to create, update, and manage chatflows via the API.
+
+### Setup
+
+1. Create a `.env` file in the project root (or in your working directory):
+
+```bash
+FLOWISE_API_URL=https://your-flowise-instance.com
+FLOWISE_API_KEY=your-api-key-here
+```
+
+2. Get your API key from Flowise:
+   - Open your Flowise instance
+   - Go to **Settings** → **API Keys**
+   - Create a new API key or copy an existing one
+
+3. The MCP server will automatically load credentials from `.env` when it starts.
+
+### Usage
+
+Once configured, you can ask Claude:
+
+- "Test the connection to my Flowise instance"
+- "List all my chatflows"
+- "Create a new chatflow with this design and deploy it"
+- "Update the existing chatflow with these changes"
+
+> **Note:** The Flowise API currently only supports **chatflows**. Agentflows must be created through the Flowise UI by importing the generated JSON.
+
+### Security Notes
+
+- Never commit your `.env` file to version control
+- The `.env` file is already in `.gitignore`
+- API keys should be kept private and rotated periodically
+
+## Alternative Installation Methods
+
+### Global Install (npm)
+
+```bash
+npm install -g ds-mcp-flowise
+```
+
+Then in your MCP config:
+
+**Windows:**
+```json
+{
+  "mcpServers": {
+    "flowise": {
+      "command": "cmd",
+      "args": ["/c", "ds-mcp-flowise"]
+    }
+  }
+}
+```
+
+**Mac / Linux:**
+```json
+{
+  "mcpServers": {
+    "flowise": {
+      "command": "ds-mcp-flowise"
+    }
+  }
+}
+```
+
+### Build from Source
+
+```bash
+git clone https://github.com/dtsoden/DS-MCP-FLOWISE.git
+cd DS-MCP-FLOWISE
+npm install
+npm run build
+```
+
+Then in your MCP config:
+
+**Windows:**
+```json
+{
+  "mcpServers": {
+    "flowise": {
+      "command": "node",
+      "args": ["C:\\path\\to\\DS-MCP-FLOWISE\\dist\\index.js"]
+    }
+  }
+}
+```
+
+**Mac / Linux:**
+```json
+{
+  "mcpServers": {
+    "flowise": {
+      "command": "node",
+      "args": ["/path/to/DS-MCP-FLOWISE/dist/index.js"]
+    }
+  }
+}
+```
+
+## Available Tools
+
+### Node Discovery
+
+| Tool | Description |
+|------|-------------|
+| `list_categories` | List all 26 node categories with counts |
+| `list_nodes` | List nodes, optionally filtered by category |
+| `get_node_schema` | Get detailed schema for a node including all inputs |
+| `search_nodes` | Search nodes by keyword |
+| `find_compatible_nodes` | Find nodes that can connect to a given node |
+
+### Template Library
+
+| Tool | Description |
+|------|-------------|
+| `list_templates` | List 64 marketplace templates by type |
+| `get_template` | Get complete template with nodes and edges |
+
+### Flow Building
+
+| Tool | Description |
+|------|-------------|
+| `validate_flow` | Validate a flow's nodes and connections |
+| `generate_flow_skeleton` | Generate a starting flow for common use cases |
+
+### Flowise API (requires .env configuration)
+
+| Tool | Description |
+|------|-------------|
+| `flowise_test_connection` | Test connection to your Flowise instance |
+| `flowise_list_chatflows` | List all chatflows in your instance |
+| `flowise_get_chatflow` | Get details of a specific chatflow |
+| `flowise_create_chatflow` | Create and deploy a new chatflow |
+| `flowise_update_chatflow` | Update an existing chatflow |
+| `flowise_delete_chatflow` | Delete a chatflow |
+
+> **Note:** The Flowise API currently only supports **chatflows**. Agentflows must be created through the Flowise UI by importing the generated JSON.
+
+## Example Prompts
+
+**Important:** You must explicitly tell Claude to use the flowise MCP tools, otherwise it may try to build something from scratch instead of using Flowise nodes.
+
+Once connected, try asking Claude:
+
+- "**Use the flowise tools** to search for Ollama nodes and build me a chatflow JSON"
+- "**Using the flowise MCP**, what vector stores does Flowise support?"
+- "**Use flowise tools** to build a simple chatbot with memory and give me the JSON to import"
+- "**With the flowise MCP**, create a RAG flow using Pinecone and OpenAI embeddings"
+- "**Use the flowise search** to find nodes that can connect to a ConversationChain"
+- "**Using flowise tools**, build me an agent that can search the web and query a database"
+
+The key is to include phrases like:
+- "use the flowise tools"
+- "using the flowise MCP"
+- "with the flowise MCP"
+- "use flowise search"
+
+## What's Included
+
+- **311 Flowise nodes** with full schemas
+- **1,915 input parameters** documented
+- **26 node categories** (Chat Models, Vector Stores, Tools, etc.)
+- **64 marketplace templates** as examples
+- **SQLite database** for fast queries
+
+## Updating the Node Database
+
+To update with the latest Flowise nodes:
+
+```bash
+git clone https://github.com/dtsoden/DS-MCP-FLOWISE.git
+cd DS-MCP-FLOWISE
+npm install
+
+# This clones Flowise and extracts all node definitions
+npm run extract
+
+# This rebuilds the SQLite database
+npm run prepare-db
+
+npm run build
+```
+
+## Credits
+
+- [Flowise](https://github.com/FlowiseAI/Flowise) - The no-code LLM orchestration platform
+- Inspired by [n8n-mcp](https://github.com/czlonkowski/n8n-mcp) architecture
+
+## License
+
+MIT
+
+---
+*Auto-published to npm via GitHub Actions*

package/dist/index.js

@@ -386,10 +386,10 @@ const TOOLS = [
 ];
 // === TOOL IMPLEMENTATIONS ===
 function listCategories() {
-    const categories = query(`
-    SELECT name, node_count
-    FROM categories
-    ORDER BY name
+    const categories = query(`
+    SELECT name, node_count
+    FROM categories
+    ORDER BY name
   `);
     return JSON.stringify(categories, null, 2);
 }
@@ -418,11 +418,11 @@ function getNodeSchema(name) {
     if (!node) {
         return JSON.stringify({ error: `Node "${name}" not found` });
     }
-    const inputs = query(`
-    SELECT input_name, input_label, input_type, description, optional, default_value, options, additional_params
-    FROM node_inputs
-    WHERE node_name = ?
-    ORDER BY additional_params, input_label
+    const inputs = query(`
+    SELECT input_name, input_label, input_type, description, optional, default_value, options, additional_params
+    FROM node_inputs
+    WHERE node_name = ?
+    ORDER BY additional_params, input_label
   `, [name]);
     const schema = {
         ...node,
@@ -443,11 +443,11 @@ function getNodeSchema(name) {
 }
 function searchNodes(searchQuery, limit = 10) {
     const searchTerm = `%${searchQuery}%`;
-    const results = query(`
-    SELECT name, label, category, description
-    FROM nodes
-    WHERE name LIKE ? OR label LIKE ? OR description LIKE ? OR category LIKE ?
-    LIMIT ?
+    const results = query(`
+    SELECT name, label, category, description
+    FROM nodes
+    WHERE name LIKE ? OR label LIKE ? OR description LIKE ? OR category LIKE ?
+    LIMIT ?
   `, [searchTerm, searchTerm, searchTerm, searchTerm, limit]);
     return JSON.stringify(results, null, 2);
 }
@@ -507,10 +507,10 @@ function findCompatibleNodes(nodeName, direction = 'inputs') {
         return JSON.stringify({ compatible_inputs: compatible }, null, 2);
     }
     else {
-        const allInputs = query(`
-      SELECT DISTINCT n.name, n.label, n.category, ni.input_type
-      FROM nodes n
-      JOIN node_inputs ni ON n.name = ni.node_name
+        const allInputs = query(`
+      SELECT DISTINCT n.name, n.label, n.category, ni.input_type
+      FROM nodes n
+      JOIN node_inputs ni ON n.name = ni.node_name
     `);
         const compatible = allInputs.filter((i) => baseClasses.includes(i.input_type)).map((i) => ({
             name: i.name,
@@ -881,13 +881,13 @@ server.setRequestHandler(GetPromptRequestSchema, async (request) => {
                     role: 'user',
                     content: {
                         type: 'text',
-                        text: `Help me build a Flowise chatflow that: ${args?.description}
-
-First, use list_categories to see available node types.
-Then, use search_nodes or list_nodes to find relevant components.
-Use get_node_schema to understand how to configure each node.
-Check list_templates for similar examples.
-Use validate_flow to verify the flow is valid.
+                        text: `Help me build a Flowise chatflow that: ${args?.description}
+
+First, use list_categories to see available node types.
+Then, use search_nodes or list_nodes to find relevant components.
+Use get_node_schema to understand how to configure each node.
+Check list_templates for similar examples.
+Use validate_flow to verify the flow is valid.
 Finally, use flowise_create_chatflow to push it directly to Flowise!`,
                     },
                 },
@@ -901,13 +901,13 @@ Finally, use flowise_create_chatflow to push it directly to Flowise!`,
                     role: 'user',
                     content: {
                         type: 'text',
-                        text: `Help me build a Flowise agent flow that: ${args?.description}
-
-Agent flows use nodes from the "Agent Flows" category.
-Use list_nodes with category="Agent Flows" to see available nodes.
-Use list_templates with type="agentflowv2" for examples.
-Key nodes include: startAgentflow, agentAgentflow, llmAgentflow, toolAgentflow, conditionAgentflow.
-Use get_node_schema to understand each node's configuration.
+                        text: `Help me build a Flowise agent flow that: ${args?.description}
+
+Agent flows use nodes from the "Agent Flows" category.
+Use list_nodes with category="Agent Flows" to see available nodes.
+Use list_templates with type="agentflowv2" for examples.
+Key nodes include: startAgentflow, agentAgentflow, llmAgentflow, toolAgentflow, conditionAgentflow.
+Use get_node_schema to understand each node's configuration.
 Finally, use flowise_create_chatflow to push it directly to Flowise!`,
                     },
                 },