Quick Start Guide
This guide will help you get started with Glance in just a few minutes.
Step 1: Start Glance
Run Glance from your terminal:
glanceYou should see output similar to:
______ _
/ ____/| |
| | __ | | ____ _ _ __ ____ ___
| | |_ || | / _' | || '_ \ / __|/ _ \
| |__| || || (_| | || | | || (__| __/
\______|_| \__,_||_||_| |_| \___|\___|
Let Your AI Understand Every Request at a Glance.
[✓] Proxy server running on localhost:15500
[✓] MCP server (SSE) running on localhost:15502/mcp
[✓] API server running on localhost:15501
[✓] Dashboard available at http://localhost:15501By default, Glance runs on these ports:
- Dashboard:
http://localhost:15501- Web UI for viewing and managing traffic - Proxy:
http://localhost:15500- The actual MITM proxy - MCP Server:
http://localhost:15502- Model Context Protocol server for AI agents
Step 2: Configure Your Client
You need to configure your application or environment to route traffic through Glance's proxy.
Option 1: One-liner Setup (Terminal)
The easiest way to get started is using the one-liner setup command:
eval "$(curl -s http://localhost:15501/setup)"This automatically sets up your current terminal session to use Glance as a proxy.
Option 2: Manual Environment Variables
Set the proxy environment variables manually:
export HTTP_PROXY=http://localhost:15500
export HTTPS_PROXY=http://localhost:15500For a persistent setup, add these lines to your shell profile (~/.bashrc, ~/.zshrc, etc.).
Option 3: Application-Specific Configuration
See the Client Configuration Guide for specific instructions for:
Step 3: View Traffic in the Dashboard
- Open your browser and navigate to
http://localhost:15501 - You'll see the Glance dashboard with a live view of captured traffic
- Make some HTTP/HTTPS requests from your configured client
- Watch the requests appear in real-time!
Step 4: Inspect Request Details
Click on any request in the traffic list to view:
- Request Headers: All HTTP headers sent with the request
- Request Body: Formatted JSON, XML, or raw body content
- Response Headers: All HTTP headers in the response
- Response Body: Formatted response content
- Timing Information: Request duration and timestamps
- cURL Export: Copy the request as a
curlcommand
Step 5: Try Creating a Mock
Let's create a simple mock to return a custom response:
- In the dashboard, click the "Rules" tab
- Click "Add Mock"
- Configure the mock:
- URL Pattern:
https://api.example.com/users - Method:
GET - Status Code:
200 - Response Body:json
{ "users": [ {"id": 1, "name": "Alice"}, {"id": 2, "name": "Bob"} ] }
- URL Pattern:
- Click "Save"
- Now any request to
https://api.example.com/userswill return your mocked response!
Step 6 (Optional): Test with cURL
Test your setup with a simple cURL command:
curl -x http://localhost:15500 https://api.github.com/usersYou should see the GitHub API response, and the request will appear in the Glance dashboard.
What's Next?
Now that you have Glance running, explore these features:
- 🔍 Traffic Inspection - Learn about advanced inspection features
- 🛠️ Mocking & Breakpoints - Master the rule engine
- 📝 Scenario Recording - Record and replay request sequences
- 🤖 MCP Integration - Connect with AI agents like Claude Desktop
Common Issues
HTTPS Certificate Errors
If you see certificate errors, you need to trust Glance's CA certificate. See the Troubleshooting Guide for instructions.
Port Already in Use
If the default ports are already in use, you can start Glance on different ports:
glance --proxy-port 8080 --dashboard-port 8081 --mcp-port 8082Can't See Any Traffic
Make sure your client is properly configured to use the proxy. Check the Client Configuration Guide for your specific environment.
Need Help?
- 📖 Read the FAQ
- 🐛 Check the Troubleshooting Guide
- 💬 Open an issue on GitHub