@docrouter/mcp 0.1.0

package/README.md

@@ -0,0 +1,207 @@
+# DocRouter MCP Server (TypeScript)
+
+A TypeScript-based Model Context Protocol (MCP) server for DocRouter API integration.
+
+## Overview
+
+This MCP server provides AI applications with access to DocRouter's document processing capabilities, including:
+
+- Document management and retrieval
+- OCR text extraction
+- AI-powered data extraction using prompts
+- Tag and prompt management
+- Search functionality
+
+## Installation
+
+```bash
+npm install
+```
+
+## Configuration
+
+The server can be configured using environment variables or command line arguments:
+
+### Environment Variables
+
+- `DOCROUTER_API_URL` - DocRouter API base URL (default: https://app.docrouter.ai/fastapi)
+- `DOCROUTER_ORG_ID` - Your DocRouter organization ID (required)
+- `DOCROUTER_ORG_API_TOKEN` - Your DocRouter organization API token (required)
+
+### Command Line Arguments
+
+- `--url` - DocRouter API base URL
+- `--org-id` - DocRouter organization ID
+- `--org-token` - DocRouter organization API token
+- `--timeout` - Request timeout in milliseconds (default: 30000)
+- `--retries` - Number of retry attempts (default: 3)
+
+## Usage
+
+### Running the Server
+
+```bash
+# Using environment variables
+npm start
+
+# Using command line arguments
+npm start -- --org-id your-org-id --org-token your-token
+
+# Development mode
+npm run start:dev
+```
+
+### Building
+
+```bash
+npm run build
+```
+
+## Available Tools
+
+### Document Tools
+- `get_docrouter_documents()` - List all documents
+- `get_docrouter_document(documentId)` - Get document by ID
+- `get_docrouter_document_ocr(documentId)` - Get raw OCR text for a document
+- `get_docrouter_document_ocr_metadata(documentId)` - Get OCR metadata for a document
+- `get_docrouter_extraction(documentId, promptRevId)` - Get extraction results
+
+### Tag Tools
+- `get_docrouter_tags()` - List all tags
+- `get_docrouter_tag(tagId)` - Get tag by ID
+
+### Prompt Tools
+- `get_docrouter_prompts()` - List all prompts
+- `get_docrouter_prompt(promptRevId)` - Get prompt by ID
+
+### Search and Extraction Tools
+- `search_docrouter_documents(query, tagIds)` - Search documents by name
+- `search_docrouter_prompts(query)` - Search prompts by name or content
+- `search_docrouter_tags(query)` - Search tags by name or description
+- `run_docrouter_extraction(documentId, promptRevId, force)` - Run AI extraction on a document
+
+### Helper Tools
+- `docrouter_help()` - Get help information about using the DocRouter API
+- `docrouter_document_analysis_guide(documentId)` - Generate a guide to analyze a specific document
+
+## Example Workflows
+
+### 1. Find and Analyze Documents
+
+```typescript
+// Search for invoice documents
+const documents = await search_docrouter_documents("invoice");
+
+// Get OCR text for the first document
+const ocrText = await get_docrouter_document_ocr(documents.documents[0].id);
+
+// Find available prompts
+const prompts = await get_docrouter_prompts();
+
+// Run extraction with a specific prompt
+const extraction = await run_docrouter_extraction(
+  documents.documents[0].id, 
+  prompts.prompts[0].id
+);
+```
+
+### 2. Document Analysis Guide
+
+```typescript
+// Get a step-by-step guide for analyzing a document
+const guide = await docrouter_document_analysis_guide("document-id");
+```
+
+## Integration with AI Applications
+
+### Claude Desktop
+
+Add to your `claude_desktop_config.json`:
+
+```json
+{
+  "mcpServers": {
+    "docrouter": {
+      "command": "node",
+      "args": ["/path/to/docrouter-mcp/dist/index.js"],
+      "env": {
+        "DOCROUTER_ORG_ID": "your-org-id",
+        "DOCROUTER_ORG_API_TOKEN": "your-token"
+      }
+    }
+  }
+}
+```
+
+### Cursor
+
+Create `.cursor/mcp.json` in your project root:
+
+```json
+{
+  "mcpServers": {
+    "docrouter": {
+      "command": "node",
+      "args": ["/path/to/docrouter-mcp/dist/index.js"],
+      "env": {
+        "DOCROUTER_ORG_ID": "your-org-id",
+        "DOCROUTER_ORG_API_TOKEN": "your-token"
+      }
+    }
+  }
+}
+```
+
+### Claude Code
+
+Create a `.mcp.conf` file in your project root:
+
+```json
+{
+  "mcpServers": {
+    "docrouter": {
+      "command": "node",
+      "args": ["/path/to/docrouter-mcp/dist/index.js"],
+      "env": {
+        "DOCROUTER_ORG_ID": "your-org-id",
+        "DOCROUTER_ORG_API_TOKEN": "your-token"
+      }
+    }
+  }
+}
+```
+
+## Development
+
+### Prerequisites
+
+- Node.js 18+
+- TypeScript 5+
+- Access to DocRouter API
+
+### Scripts
+
+- `npm run build` - Build the project
+- `npm run dev` - Build in watch mode
+- `npm run start:dev` - Run in development mode
+- `npm run lint` - Run ESLint
+- `npm run type-check` - Run TypeScript type checking
+
+### Project Structure
+
+```
+src/
+├── index.ts          # Main MCP server implementation
+package.json          # Package configuration
+tsconfig.json         # TypeScript configuration
+tsup.config.ts        # Build configuration
+README.md             # This file
+```
+
+## Error Handling
+
+The server includes comprehensive error handling and will return structured error responses for debugging. All errors are logged to stderr to avoid interfering with MCP communication.
+
+## License
+
+MIT