WorkplaceSearchAgent MCP Server
Enables AI agents to search, retrieve, and summarize content from workplace tools including Google Drive, Notion, Slack, and Confluence through secure Model Context Protocol.
README
AI-Powered Workplace Search - WorkplaceSearchAgent MCP Server[Hackathon: https://www.hackerearth.com/challenges/hackathon/mcp-hackathon]
Project Description
A comprehensive Model Context Protocol (MCP) server that enables AI agents to search, retrieve, and summarize content from workplace tools including Google Drive, Notion, Slack, and Confluence. This remote MCP server allows Claude Desktop to seamlessly interact with your workplace documents through secure, authenticated connections while maintaining proper user permissions and access scopes.
Team Information
Team Name: Team Lewok
Team Members:
- Chetan Mali
- Abhijeet Rajput
Hackathon Theme
Theme 2: Secure Agent-API Interface Layer
This project addresses the critical challenge of creating a reliable, secure interface layer for autonomous agents to interact with real-world APIs. As agents transition from research to production environments, they require robust identity, scoping, and trust frameworks to operate autonomously at scale. Our MCP server serves as a foundational building block in this new software architecture, enabling agents to safely invoke tools and automate operations with integrated Descope authentication and Cequence MCP proxy security.
YouTube Link
https://youtu.be/SqHEs32dtBQ?si=rBs-IuEbRcR0giB8
What We Built
Core Features
- š Universal Workplace Search: Unified search across Google Drive, Notion, Slack, and Confluence
- š Document Content Retrieval: Read full content of specific documents with proper formatting
- š¤ AI-Powered Summarization: Extract key points and generate summaries from retrieved documents
- š Secure Authentication: User-specific access control using Descope authentication framework
- š”ļø Proxy Security: All agent interactions secured through Cequence MCP proxy
- ā° Recent Updates Tracking: Pull and display recently modified documents across platforms
- šÆ Smart Query Processing: Natural language understanding for complex search requests
Available Tools for Claude Desktop
1. Search Google Drive
- Functionality: Natural language search across all Google Drive documents
- Usage: "Search my Google Drive for quarterly reports" or "Find documents about project timeline"
- Returns: Structured results with document metadata, summaries, and direct links
2. Get Document Content
- Functionality: Retrieve full content of specific documents by ID
- Usage: "Read the content of document ID 1BxY2zW3vU4sR5qP6oN7mL8kJ9hG"
- Returns: Complete document content with formatting and metadata
3. Recent Updates Monitor
- Functionality: Display recently modified or created documents
- Usage: "Show me documents updated in the last 7 days"
- Returns: Chronological list of recent changes across connected platforms
4. Document Summarization
- Functionality: Generate intelligent summaries of workplace documents
- Usage: "Summarize the Q4 budget document" or "Give me key points from the project proposal"
- Returns: Structured summaries with key insights and action items
Architecture Components
- MCP Server Core: Implements Model Context Protocol for seamless agent communication
- Authentication Layer: Descope integration for secure, token-based user authentication
- Multi-Platform Adapters: Individual adapters for Google Drive, Notion, Slack, and Confluence
- Security Proxy: Cequence MCP proxy ensuring secure agent interactions
- Mock Development Services: Full mock implementations for testing and development
How to Run
Prerequisites
- Python 3.8+
- Active internet connection for remote server access
- Google Drive account with API access
- Descope authentication credentials
- Claude Desktop application
Quick Setup (2 minutes)
Step 1: Start the MCP Server
# Start the server on port 8000
python start_server.py
Server will be accessible at http://localhost:8000
Step 2: Configure Claude Desktop
Option A: Automatic Configuration (Recommended)
# Run the MCP client configuration
python claude_mcp_client.py
Option B: Manual Configuration
-
Locate Claude Desktop config file:
- Windows:
%APPDATA%\Claude\claude_desktop_config.json - macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Linux:
~/.config/claude/claude_desktop_config.json
- Windows:
-
Add MCP server configuration:
{
"mcpServers": {
"workplace-search-mcp": {
"transport": {
"type": "http",
"url": "http://localhost:8000/mcp"
}
}
}
}
- Restart Claude Desktop
Step 3: Google Drive OAuth Setup
# Run OAuth setup utility
python test.py
# Follow the interactive setup instructions
Step 4: Test the Connection
- Open Claude Desktop
- Start a new conversation
- Ask: "Search my Google Drive for project documents"
- Claude should now access your workplace tools!
Development Setup
# Clone and install dependencies
git clone <repository-url>
cd WorkSpaceAgent_New
pip install -r requirements.txt
# Configure environment
cp .env.example .env
# Edit .env with your API keys and credentials
# Run tests
python -m pytest tests/
Tech Stack
Required Technologies (Core)
- Python 3.8+: Primary development language
- Model Context Protocol (MCP): Standard for agent-tool communication
- Descope: Authentication and identity management platform
- Cequence: MCP security proxy for agent interactions
Workplace API Integrations
- Google Drive API: Document search, retrieval, and metadata access
- Notion API: Workspace content and database access
- Slack API: Message search and channel content retrieval
- Confluence API: Knowledge base and wiki integration
Development Libraries
- asyncio: Asynchronous programming for concurrent operations
- dotenv: Secure environment variable management
- google-auth-oauthlib: OAuth2 authentication flow implementation
- googleapiclient: Official Google API client library
- httpx: Modern HTTP client for API communications
- pydantic: Data validation and settings management
Security and Authentication Stack
- OAuth 2.0: Industry-standard authorization framework
- JWT Tokens: Secure session and identity management
- Scoped Permissions: Granular, user-specific access control
- HTTPS/TLS: Encrypted data transmission
- Proxy Layer: Cequence MCP security proxy for agent protection
Project Structure
WorkSpaceAgent_New/
āāā app/
ā āāā auth/
ā ā āāā descope_auth.py # Descope authentication handler
| | |āā cequence_gateway.py # Cequence Gateway Handler
| | āāā security.py # Security Layer
ā āāā adapters/
ā ā āāā google_drive_adapter.py # Google Drive integration
ā āāā utils/
| āāā logger.py # Logger
āāā start_server.py # Server startup script
āāā claude_mcp_client.py # Claude Desktop configuration
āāā test.py # OAuth setup and testing
āāā requirements.txt # Python dependencies
āāā .env.example # Environment template
āāā USER_GUIDE.md # End-user documentation
Usage Examples
Basic Workplace Search
User: "Search my Google Drive for budget documents"
Claude: I'll search your Google Drive for budget documents.
Found 3 documents:
1. **Q4 Budget Planning**
Author: Finance Team
Last Modified: 2025-09-01T14:30:00Z
Summary: Updated quarterly budget allocations for Q4...
Document ID: 1BxY2zW3vU4sR5qP6oN7mL8kJ9hG
Smart Content Analysis
User: "Find all documents related to the marketing campaign and summarize key points"
Claude: Found 5 marketing campaign documents. Here are the key insights:
**Campaign Strategy Overview:**
- Target audience: 25-45 professionals
- Budget allocation: $50K digital, $30K traditional
- Timeline: Q4 2025 launch
**Action Items:**
- Creative assets due Sept 15
- Media buy confirmation needed
- Analytics dashboard setup required
Privacy & Security
Data Access Model
- Read-only permissions: Server can only read existing files, no modification capabilities
- Scope-limited access: Only accesses files user has explicit permission to view
- No persistent storage: Document content is not cached or stored on server
- Audit trail: All access logged for security compliance
Authentication Security
- OAuth 2.0 standard: Industry-standard authorization without credential sharing
- Token-based access: Secure, time-limited access tokens
- User consent: Explicit user authorization for each connected service
- Revocable permissions: Users can revoke access at any time
Data Transmission
- HTTPS encryption: All communications use TLS encryption
- No data logging: File content not permanently logged or stored
- Minimal data transfer: Only requested content and metadata transmitted
- Secure proxy: Cequence MCP proxy adds additional security layer
Innovation Highlights
This project represents a significant advancement in secure agent-API interaction patterns:
- First-class MCP Implementation: Native Model Context Protocol support for seamless agent integration
- Multi-platform Abstraction: Unified interface across diverse workplace tools
- Security-first Design: Built-in authentication, authorization, and audit capabilities
- Production-ready Architecture: Scalable, maintainable codebase with comprehensive testing
- Developer Experience: Simple setup, clear documentation, and robust error handling
Future Roadmap
- Additional Platform Support: Teams, Asana, Jira integration
- Advanced AI Features: Semantic search, content recommendations
- Enterprise Features: SSO integration, advanced audit logging
- Performance Optimization: Caching, batch operations, real-time updates
- Mobile Support: Mobile app integration and responsive design
*This project demonstrates how AI agents can securely and efficiently access workplace knowledge, providing a robust foundation for the next generation of autonomous workplace assistance while maintaining the highest standards of security and user
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.