Hedy MCP Server: Developer Guide to 18 Tools and API Access
Technical companion to the Hedy MCP Server launch. Detailed reference for all 18 specialized tools, OAuth 2.1 authentication, MCP protocol compatibility, real-world workflow patterns, and the January 2026 write-capabilities update.
https://api.hedy.bot/mcp on the v1.4.0 REST API. For the original launch announcement, see the overview post.
Developer Guide to the Hedy MCP Server
This is the technical companion to our original MCP launch announcement. It documents the full tool catalog, authentication model, and workflow patterns for developers connecting Claude or other MCP-compatible clients to their Hedy meeting data through your AI meeting assistant with 18 available tools and enhanced API capabilities.
Hedy connects directly with Claude and other AI assistants through the Model Context Protocol (MCP). Your AI tools can access, organize, and work with your meeting data without manually copying information between applications.
What This Actually Does
The MCP server gives AI assistants like Claude Desktop direct access to your Hedy data through 18 specialized tools. Here’s what becomes possible:
Session Management
AI assistants can retrieve your meeting transcripts, summaries, and insights. Ask Claude to “find the budget discussion from last week’s client meeting” and it pulls the exact session with full context.
Topic Organization
Create and manage topics programmatically. Your AI can automatically organize related meetings, build knowledge bases from conversation series, and track project evolution across multiple sessions.
Highlight Capture
Access saved highlights with their AI-generated analysis. Build study guides from lecture highlights, compile decision points from business meetings, or extract key quotes from interviews.
Task Integration
Pull to-do items directly into your workflow. AI assistants can aggregate action items across projects, identify overdue tasks, or create consolidated task lists from multiple meetings.
Technical Implementation
The MCP server runs at https://api.hedy.bot/mcp and authenticates using your Hedy API key. Configuration takes about two minutes:
- Generate an API key in Hedy’s Account Settings
- Add the MCP server URL to Claude Desktop or your preferred MCP-compatible client
- Configure authentication with your API key
- Start querying your meeting data through natural language
The integration supports both session-level queries (“What did we discuss about pricing?”) and cross-session analysis (“Show me all action items from client meetings this month”).
Privacy Architecture
The MCP integration maintains Hedy’s privacy-first design. Data remains encrypted in transit and at rest. The MCP server provides read access to your meeting data—AI assistants can retrieve and analyze information but cannot modify your original sessions or transcripts.
Authentication is user-specific. Only AI tools configured with your API key can access your data. Keys can be revoked instantly through Account Settings if needed.
Available Tools
The MCP server provides 18 specialized tools organized by function:
Session Access
- Get session details with full transcripts and metadata
- List sessions with filtering by date, topic, or type
- Search across sessions for specific content
Topic Management
- Create topics with custom contexts and instructions
- List topics with session counts and metadata
- Retrieve topic details including all related sessions
- Update topic configurations programmatically
Highlights & Tasks
- Access highlights with AI-generated insights
- List to-dos with completion status
- Filter highlights by session or topic
- Retrieve task details with deadlines and metadata
Real-World Applications
Research Analysis
Academics use the integration to synthesize findings across lecture series. Ask Claude to “compare the three theoretical frameworks discussed in weeks 4-6” and it pulls relevant content from multiple sessions, cross-references concepts, and generates comparative analysis.
Client Management
Consultants maintain comprehensive client histories. The AI can “prepare for tomorrow’s meeting with [Client] by reviewing all previous discussions about their Q4 strategy” and generates a briefing document with relevant context, outstanding commitments, and suggested talking points.
Knowledge Management
Teams build institutional knowledge. The MCP integration lets AI assistants create structured documentation from meeting series, extract best practices from retrospectives, and compile technical specifications from design discussions.
Workflow Automation
Developers integrate Hedy data into custom systems. Combine MCP access with n8n or Make.com to build automated workflows that route meeting insights to project management tools, update CRMs, or trigger follow-up actions.
Requirements & Compatibility
The MCP integration works with any MCP-compatible client. Confirmed compatible applications include:
- Claude Desktop (macOS and Windows)
- Cline
- Zed
- Custom implementations using the MCP SDK
A Hedy Pro subscription provides full API access required for MCP integration. Free tier users have limited API access suitable for testing but not production workflows.
Getting Started
Complete setup instructions are available in Hedy’s help documentation. The configuration process involves generating an API key, adding the MCP server to your AI client, and testing the connection with a simple query.
For developers building custom integrations, API documentation covers authentication methods, rate limits, and response schemas. The MCP server supports standard MCP protocol features including tool discovery and capability negotiation.
What This Enables
The MCP integration transforms how professionals use meeting intelligence. Instead of manually searching through past sessions or copying information into documents, AI assistants handle the heavy lifting—finding relevant context, synthesizing information across conversations, and presenting actionable insights.
This matters because professional work increasingly happens through conversations. The knowledge captured in meetings, calls, and discussions represents significant organizational intelligence. The MCP integration makes that intelligence accessible to the AI tools you already use for analysis, writing, and decision-making.
For organizations, this means meeting data becomes a queryable knowledge base. For individuals, it means every conversation you’ve had with Hedy becomes available to assist with current work—no manual organization required.
Technical Foundation
The implementation follows MCP specification version 1.0, ensuring compatibility with the growing ecosystem of MCP-enabled applications. The server architecture supports concurrent requests, implements efficient caching for frequently accessed data, and includes automatic retry logic for transient failures.
Response times typically range from 100-500ms for session retrieval and 200-1000ms for complex cross-session queries, depending on data volume and analysis depth.
Future Development
The current implementation provides comprehensive read access to Hedy data. Future updates will add write capabilities, enabling AI assistants to create highlights during analysis, add custom tags to sessions, and append notes to meetings programmatically.
Integration with additional protocols beyond MCP is under evaluation based on ecosystem adoption and user requests.
Getting Help
Documentation is available in Hedy’s help center, including setup guides for popular AI clients, troubleshooting common configuration issues, and example queries for different use cases.
The Hedy Slack community includes an #mcp-integration channel where users share configurations, discuss workflows, and provide mutual support for technical implementation questions.
The MCP integration is available now for Hedy Pro users. Generate your API key in Account Settings to begin connecting your AI assistants to your meeting intelligence.
Frequently Asked Questions
How many tools does the Hedy MCP Server expose?
18 specialized tools organized into four groups: Session Access (transcripts, search, listings), Topic Management (create/update/list topics with custom contexts), Highlights & Tasks (AI insights, completion status, deadlines), and Webhooks/Session Contexts (write-capable as of January 2026).
Does the Hedy MCP server support write operations?
Yes, as of January 2026. AI assistants can create and manage topics and session contexts, configure webhooks, and update topic configurations. Read-only access remains the default, but write-capable tools are exposed where the operation makes sense.
How does Hedy MCP authentication work?
The Hedy MCP server uses OAuth 2.1. Your AI client opens a browser window for sign-in and permission review; an access token is issued once you authorize. No static API keys to rotate. Tokens can be revoked from Hedy Account Settings.
What’s the Hedy MCP server URL?
https://api.hedy.bot/mcp for US accounts. EU data-residency accounts use https://eu-api.hedy.bot/mcp. Add the URL to your MCP-compatible client (Claude Desktop, Cline, Zed, Cursor, Claude Code) and authenticate via OAuth.
What are the MCP server response times and rate limits?
Session retrieval typically completes in 100–500ms; cross-session queries in 200–1000ms depending on data volume and analysis depth. The server supports concurrent requests with efficient caching for frequently accessed data and automatic retry logic for transient failures.
