heizaheiza

Charles Proxy MCP server with live capture and structured traffic analysis

29
3
100% credibility
Found Mar 07, 2026 at 28 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A tool that connects the Charles web debugging proxy to AI assistants for real-time and historical network traffic inspection with filtering, grouping, and automatic privacy protection.

How It Works

1
πŸ” Discover the traffic helper

You hear about a handy tool that lets your AI assistant safely peek at web connections from your browsing or app.

2
πŸ’» Set up the capture app

Download and start the Charles app on your computer to begin recording what websites your browser talks to.

3
πŸ” Enable easy sharing

In the app, turn on web sharing with a simple username and password so your tools can access the recordings.

4
πŸ€– Connect to your AI

Link this friendly bridge to your AI chat helper, like Claude, with a quick setup so it can read the captures.

5
▢️ Start watching traffic

Tell it to begin live recording or check past sessions, and your AI starts seeing the web conversations.

6
πŸ“Š Ask for smart reviews

Chat with your AI to filter, group, or dive into details of the traffic, with sensitive info automatically hidden.

πŸŽ‰ Unlock clear insights

Your AI delivers easy summaries, spots issues, and helps debug web problems without overwhelming details or risks.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 28 to 29 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is Charles-mcp?

Charles-mcp is a Python MCP server that plugs into Charles Proxy for agent-driven network inspection. It streams live traffic increments from active recordings, runs structured analysis on live or saved sessions, and filters noise like images, fonts, and tunnels while redacting tokens, cookies, and auth headers by default. Users get summary-first outputs via tools like `start_live_capture`, `group_capture_analysis`, and `get_traffic_entry_detail` to drill down without token waste.

Why is it gaining traction?

It beats raw Charles Proxy APIs or mitmproxy by enabling true realtime polling agents can read mid-recording, with presets like "api_focus" prioritizing APIs over static assets. Token-smart grouping by host, path, or status spots issues fast, and recoverable stops prevent lost dataβ€”ideal for Claude CLI or Antigravity users. Bilingual docs cover Charles proxy chrome, firefox, ios, and android setups seamlessly.

Who should use this?

Backend devs debugging API flows with Charles Proxy 5 via AI agents, mobile engineers capturing Charles proxy android/ios traffic for structured review, or QA teams analyzing Charles proxy certificate errors in live browser sessions through MCP clients like Codex CLI.

Verdict

At 19 stars and 1.0% credibility, this beta tool shows solid docs and pytest coverage but needs more adoption for production trust. Use it now if MCP + Charles fits your workflow; it's a smart enhancer for agent debugging.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.