SETUP_CURSOR.md•6.85 kB
# Using Resume MCP Server with Cursor IDE
## Overview
Cursor doesn't natively support MCP servers like Claude Desktop does, but you can integrate the resume server functionality in several ways.
## Option 1: Direct Python Integration (Recommended)
### Setup
1. Open your project in Cursor
2. Open the terminal in Cursor (Cmd+` or View → Terminal)
3. Install dependencies:
```bash
pip install "mcp[cli]"
```
### Usage in Cursor
Create a helper file `cursor_helper.py`:
```python
# cursor_helper.py
import sys
sys.path.append('/path/to/rishi-resume-mcp')
from server import (
get_ai_ml_experience,
search_experience,
get_metrics_and_impact,
get_company_details
)
# Now you can use these functions directly in Cursor's AI chat
# Example: "Using get_ai_ml_experience(), show me Rishi's AI work"
```
Then in Cursor's AI chat (Cmd+K), you can:
- Reference the functions: `@cursor_helper.py`
- Ask: "Run get_ai_ml_experience() and summarize the results"
## Option 2: Custom Cursor Commands
### Create Custom Commands
1. Create `.cursor/commands.json` in your project:
```json
{
"commands": [
{
"name": "Resume: AI Experience",
"command": "python -c \"from server import get_ai_ml_experience; import json; print(json.dumps(get_ai_ml_experience(), indent=2))\"",
"cwd": "/path/to/rishi-resume-mcp"
},
{
"name": "Resume: Search",
"command": "python -c \"from server import search_experience; import json; import sys; query = ' '.join(sys.argv[1:]); print(json.dumps(search_experience(query), indent=2))\"",
"cwd": "/path/to/rishi-resume-mcp",
"args": true
},
{
"name": "Resume: Metrics",
"command": "python -c \"from server import get_metrics_and_impact; import json; print(json.dumps(get_metrics_and_impact(), indent=2))\"",
"cwd": "/path/to/rishi-resume-mcp"
}
]
}
```
2. Run via Command Palette (Cmd+Shift+P → "Run Custom Command")
## Option 3: Interactive Terminal Session
### Quick Access Script
Create `cursor_resume.py`:
```python
#!/usr/bin/env python3
"""
Cursor IDE Resume Query Interface
"""
import json
import sys
import os
# Add resume server to path
sys.path.insert(0, '/path/to/rishi-resume-mcp')
from server import *
def query(question):
"""Quick query function for Cursor terminal"""
question_lower = question.lower()
if 'ai' in question_lower or 'ml' in question_lower:
return json.dumps(get_ai_ml_experience(), indent=2)
elif 'metric' in question_lower or 'impact' in question_lower:
return json.dumps(get_metrics_and_impact(), indent=2)
elif 'compan' in question_lower:
companies = ["Justworks", "Stash", "Casper", "Peet's"]
results = {}
for company in companies:
results[company] = get_company_details(company)
return json.dumps(results, indent=2)
else:
return json.dumps(search_experience(question), indent=2)
# If run directly, take command line argument
if __name__ == "__main__":
if len(sys.argv) > 1:
result = query(" ".join(sys.argv[1:]))
print(result)
else:
print("Usage: python cursor_resume.py <question>")
print("Example: python cursor_resume.py 'AI experience'")
```
Then in Cursor's terminal:
```bash
python cursor_resume.py "What AI models has Rishi built?"
```
## Option 4: Cursor AI Context
### Add Resume Data as Context
1. Open `resume_data.json` in Cursor
2. In Cursor Chat (Cmd+K), reference it: `@resume_data.json`
3. Ask questions directly:
- "Based on @resume_data.json, what ML models has Rishi built?"
- "Analyze @resume_data.json and list revenue impact"
### Create Context Snippets
Create `.cursor/context/resume_queries.md`:
```markdown
# Resume Query Templates
## AI/ML Experience
Look in resume_data.json → ai_experience → models_built, tools_used, initiatives_led
## Revenue Impact
Look in resume_data.json → key_metrics → revenue_impact
## Company Experience
Look in resume_data.json → experience → filter by company name
## Skills
Look in resume_data.json → skills → [category]
```
Reference this in Cursor chat: `@resume_queries.md`
## Option 5: VSCode/Cursor Extension (Advanced)
Create a simple extension that wraps the MCP server:
1. Create extension structure:
```
resume-mcp-extension/
├── package.json
├── extension.js
└── README.md
```
2. `package.json`:
```json
{
"name": "resume-mcp",
"version": "0.1.0",
"main": "./extension.js",
"contributes": {
"commands": [
{
"command": "resume.queryAI",
"title": "Resume: Query AI Experience"
},
{
"command": "resume.queryMetrics",
"title": "Resume: Query Business Metrics"
}
]
}
}
```
3. `extension.js`:
```javascript
const vscode = require('vscode');
const { exec } = require('child_process');
function activate(context) {
let queryAI = vscode.commands.registerCommand('resume.queryAI', () => {
exec('python /path/to/rishi-resume-mcp/test_server.py', (err, stdout) => {
vscode.window.showInformationMessage(stdout);
});
});
context.subscriptions.push(queryAI);
}
module.exports = { activate };
```
## Best Practices for Cursor
1. **Use @-mentions**: Reference files directly in Cursor chat
- `@server.py` - Reference the server code
- `@resume_data.json` - Reference the data
- `@test_server.py` - Reference test functions
2. **Create Snippets**: Save common queries as snippets
```python
# .cursor/snippets/resume_ai.py
from server import get_ai_ml_experience
print(get_ai_ml_experience())
```
3. **Terminal Integration**: Keep a terminal tab open with:
```bash
python test_server.py
```
4. **Cursor Composer**: Use Composer mode (Cmd+I) and reference:
- "Using the functions in @server.py, analyze Rishi's AI experience"
- "Run @test_server.py and show me revenue metrics"
## Example Cursor Workflow
1. Open Cursor in the resume-mcp directory
2. In Cursor Chat (Cmd+K):
```
@server.py Can you run get_ai_ml_experience() and format the results?
```
3. Or in Composer (Cmd+I):
```
Using @resume_data.json, create a summary of Rishi's platform migrations
```
## Debugging in Cursor
1. Set breakpoints in `server.py`
2. Run in debug mode:
```bash
python -m debugpy --listen 5678 --wait-for-client test_server.py
```
3. Attach Cursor's debugger to port 5678
## Quick Test
In Cursor's terminal:
```bash
# Test that imports work
python -c "from server import get_ai_ml_experience; print('✅ Server loads correctly')"
# Run interactive test
python test_server.py
```
---
**Note**: While Cursor doesn't have native MCP support like Claude Desktop, these integration methods let you effectively query and work with the resume data within the IDE.