Neura Market
Anything Analyzer logo

Anything Analyzer

Free

全能协议分析工具：浏览器抓包 + MITM 代理 + 指纹伪装 + AI 分析 + MCP Server 无缝对接 AI Agent/IDE | All-in-one protocol analysis toolkit — built-in browser capture, MITM proxy, JS hooks, fingerprint spoofing, AI analysis & M

FreeFree tier
#TypeScript
Type
Saas
Visit Website GitHub (2.2k stars)

About Anything Analyzer

A comprehensive all-in-one protocol analysis toolkit that captures network traffic from multiple sources — web browsers, desktop applications, terminal commands, scripts, and mobile/IoT devices — and leverages AI to automatically analyze and reverse-engineer protocols. Built-in browser capture via CDP, MITM proxy on port 8888, JS hook injection for intercepting encryption calls, and MCP server integration allow AI agents (Claude Desktop, Cursor, etc.) to directly use its capture and analysis capabilities. Features two-phase AI analysis (smart noise filtering then deep analysis) with five analysis modes: auto-identify, API reverse, security audit, performance analysis, and JS encryption reverse. Supports streaming output with multi-turn questioning for interactive exploration.

Key Features

Full-scenario packet capture: built-in browser (CDP), MITM proxy (port 8888), supports web, desktop apps, terminal commands, scripts, mobile/IoT
AI intelligent analysis with two-phase processing: smart noise filtering then deep analysis
5 analysis modes: auto-identify, API reverse, security audit, performance analysis, JS encryption reverse
JS Hook injection: automatically intercepts fetch, XHR, crypto.subtle, CryptoJS, SM2/3/4, and other encryption calls
Encryption code extraction: automatically extracts encryption-related code snippets from JS files
Streaming output with multi-turn questioning for real-time interactive analysis
Unified session: all requests from different sources merge into one session for comprehensive AI analysis
MCP ecosystem integration: built-in MCP Server exposing capture/analysis as tools for AI agents (Claude Desktop, Cursor, etc.) and MCP Client to access external MCP servers

Pros & Cons

Pros
  • Captures traffic from virtually any source (web, desktop, terminal, mobile, IoT) in one unified session
  • AI-powered automatic analysis significantly reduces manual packet inspection effort
  • JS hooking capability enables deep reverse engineering of client-side encryption
  • MCP integration allows AI agents and IDEs to directly consume capture and analysis capabilities
  • Open source and free to use with no hidden pricing tiers
  • Multi-turn questioning enables interactive drill-down into analysis results
Cons
  • Requires manual CA certificate installation for MITM proxy to capture HTTPS traffic
  • Depends on external LLM API (OpenAI/Anthropic) for AI analysis, requiring users to provide their own API key
  • No native mobile app; mobile traffic capture requires device proxy configuration
  • Electron-based desktop app may be resource-intensive
  • Learning curve for users unfamiliar with proxy configuration and certificate installation

Best For

Reverse engineer website APIs: analyze embedded browser traffic to generate API endpoint docs, authentication flows, and Python reproduction codeReverse engineer mobile app protocols: capture traffic via Wi-Fi proxy to discover hidden APIs and request signing logicJS encryption reverse engineering: use embedded browser with JS hook to identify encryption algorithms, restore processes, and implement Python counterpartsSecurity auditing: detect token leaks, CSRF/XSS vulnerabilities, and sensitive data exposure across hybrid traffic sourcesDebugging CLI tools: capture traffic from terminal curl/httpie to get complete request/response records with AI interpretationDebugging microservices: capture inter-service traffic via script configuration and environment variable proxy to analyze call chains and authentication flows

FAQ

How do I capture traffic from a mobile app?
Set up the MITM proxy on port 8888, install the CA certificate on your device, and configure the device's Wi-Fi proxy to point to the machine running Anything Analyzer.
What AI models are supported?
Anything Analyzer supports any LLM API compatible with OpenAI or Anthropic formats. You configure the API key in Settings → LLM.
Can I integrate Anything Analyzer with other AI agents?
Yes, the tool provides a built-in MCP Server that exposes capture and analysis capabilities as MCP tools, allowing agents like Claude Desktop and Cursor to call them directly.