Documentation Index
Fetch the complete documentation index at: https://docs.usehindsight.com/llms.txt
Use this file to discover all available pages before exploring further.
Model Context Protocol (MCP) Integration
Hindsight exposes its AI agent capabilities through the Model Context Protocol (MCP), allowing AI assistants to directly query your deal data, research competitors, and analyze pipeline metrics.
Compatible with: Claude Desktop, Claude Code, ChatGPT, Cursor, Cline, and any MCP-compatible client.
Generate an API key from Settings → API Keys in the Hindsight app before getting started.
Setup
Claude
OpenAI / ChatGPT
Direct API
One-Click Install (Preferred Method)
Download and double-click the Hindsight MCP bundle to install directly into Claude Desktop. No config file editing required.Download hindsight.mcpbWhen you open the file and click install, Claude Desktop will open the browser prompting you to login to Hindsight and select an organization to use for the MCP.
Manual Setup
If you prefer to configure manually, open your Claude Desktop config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json
- Windows:
%APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"hindsight": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://app.usehindsight.com/api/mcp", "--header", "Authorization: Bearer YOUR_API_KEY"]
}
}
}
Claude Code
Run this command in your terminal:claude mcp add hindsight --transport http https://app.usehindsight.com/api/mcp \
--header "Authorization: Bearer YOUR_API_KEY"
ChatGPT Developer Mode
ChatGPT’s MCP support is available in beta via Developer Mode on Pro, Plus, Business, Enterprise, and Education accounts.1. Enable Developer ModeGo to Settings → Apps → Advanced settings → Developer mode and toggle it on.2. Create a Hindsight app
- Open ChatGPT Apps settings
- Click Create app (only visible when Developer Mode is enabled)
- Enter
https://app.usehindsight.com/api/mcp as the MCP server URL
- Set authentication to “OAuth”
- After hitting “Create” you should be directed to a new page to login with your Hindsight account to authenticate
The app will appear under Drafts in your app settings.3. Use Hindsight in a conversationSelect Developer Mode from the composer’s Plus menu, choose the Hindsight app, and start your conversation. For best results, be explicit about which tool to use:
“Use the Hindsight app’s research_deals tool to find why we lost deals to Salesforce last quarter.”
For more details, see OpenAI’s Developer Mode documentation.HTTP / JSON-RPC
The Hindsight MCP server is a standard HTTP endpoint. Any HTTP client can call it directly using JSON-RPC 2.0.Endpoint: https://app.usehindsight.com/api/mcpRequired headers:Authorization: Bearer YOUR_API_KEY
Content-Type: application/json
Accept: application/json, text/event-stream
curl -X POST https://app.usehindsight.com/api/mcp \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}'
curl -X POST https://app.usehindsight.com/api/mcp \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{
"jsonrpc": "2.0",
"id": 2,
"method": "tools/call",
"params": {
"name": "research_competitive_intel",
"arguments": { "question": "Who are our top competitors?" }
}
}'
1. pull_schema
Fetch the schema of your deal grid — field names, types, and structure. Use this before create_dataset to understand what data is available.
2. create_dataset
Export your deal grid data as a structured JSON dataset. Returns a json_id that you pass to analyze_deal_data.
3. analyze_deal_data
Analyze quantitative deal data: win rates, trends, counts, and metrics across your pipeline. Requires a json_id from create_dataset.
Parameters:
| Parameter | Type | Description |
|---|
question | string | The data analysis question to answer |
json_id | string | Dataset ID returned by create_dataset |
1. Call pull_schema → understand your deal fields
2. Call create_dataset → get a json_id
3. Call analyze_deal_data with your question + json_id
- “What is our win rate by industry?”
- “Which competitors do we lose to most often?”
- “What’s the average deal size for closed won deals in Q1?“
4. research_competitive_intel
Research competitive intelligence using Hindsight’s knowledge base and live web search. Handles multi-step research internally.
Parameters:
| Parameter | Type | Description |
|---|
question | string | The competitive intelligence question to research |
- “What are Salesforce’s main weaknesses based on our deals?”
- “How should we position against HubSpot?”
- “What recent news is there about Gong?“
5. research_deals
Research qualitative insights from deal transcripts, emails, call recordings, and CRM notes. Handles multi-step document retrieval internally.
Parameters:
| Parameter | Type | Description |
|---|
question | string | The deal research question to investigate |
- “Why are we losing deals to Competitor X?”
- “What objections come up most often in enterprise deals?”
- “What did prospects say about our pricing in lost deals last quarter?”
How It Works
Each tool delegates to Hindsight’s AI agent pipeline. When you call research_competitive_intel or research_deals, Hindsight runs a full multi-step agent behind the scenes — searching documents, querying your knowledge base, and synthesizing a response — and returns the final answer.
For analyze_deal_data, Hindsight runs structured queries against your exported deal grid and computes the metrics you ask for.
Authentication
MCP requests authenticate using your Hindsight API key as a Bearer token. Keys are scoped to your organization — all tools operate on your org’s data only.
Generate an API key from Settings → API Keys in the Hindsight app.
