httptrafficmonitr

# HTTP Traffic Monitor A Windows desktop app that acts as a local HTTP/HTTPS man-in-the-middle proxy, giving you a Fiddler/Charles-style traffic inspector with modern WPF Fluent UI — plus an **MCP server** so AI assistants (Claude, etc.) can drive it directly.  ## What it does `HttpTrafficMonitor` sits between your machine and the internet as a local proxy (via [Titanium.Web.Proxy](https://github.com/justcoding121/Titanium-Web-Proxy)), decrypts HTTPS on the fly with a locally-generated root CA, and shows every request/response as it happens — searchable, filterable, and exportable.  ## Features ### Capture & inspect - Start/Stop system-wide HTTP/HTTPS capture with one click (installs a root CA for TLS interception, restored automatically on Stop) - Live request grid: method, URL, status, size, duration, colored badges, and a red highlight for slow requests - Quick search plus dedicated Process / Domain / Method / Status filters - Per-request detail tabs: **Request**, **Response**, **Timing** (DNS/TCP/TLS/TTFB/download breakdown), **Certificate** (chain, issuer, validity, thumbprint), **Decoded** (gzip/brotli/base64 aware), and **Exclusions**    ### Filtering Build multi-condition filters (field / operator / value, AND-ed or OR-ed) and save them as named presets for reuse.  ### Replay & compose Edit and resend any captured request, or compose a brand-new one from scratch, with side-by-side original vs. new response.  ### Compare Pick two requests and get a line-by-line diff (via [DiffPlex](https://github.com/mmanela/diffplex)) across headers and bodies. ### Alerts Rule-based alerting on status codes, response time thresholds, or custom patterns, with a live event feed.  ### Auto-Responder (mocking) Intercept matching requests (by URL/regex + method) and return a canned status code, headers, and body — or serve a file — with an optional artificial delay. Great for mocking flaky or unavailable backends during development.  ### Graphs Live requests-per-second and bandwidth charts, plus top-domains and top-processes breakdowns (via [LiveCharts2](https://livecharts.dev/)). ### Sessions & export Save/load full capture sessions, bookmark interesting requests, and export selections as **CSV**, **HAR**, **JSON**, **cURL**, or **Postman collection**. ### Light & dark themes Full Fluent Design theming via [WPF-UI](https://github.com/lepoco/wpfui).  ## MCP Server `HttpTrafficMonitor.McpServer` exposes the running app to AI assistants over the [Model Context Protocol](https://modelcontextprotocol.io) (stdio transport). The main app hosts a small local REST API (`http://localhost:18081/api/...`) via `IpcApiService`; the MCP server is a thin process that talks to that API and surfaces it as MCP tools — so an assistant can start/stop capture, inspect and filter requests, manage alerts and auto-responder rules, replay/compose requests, and export data, all without touching the UI. Tool groups (see `HttpTrafficMonitor.McpServer/Tools/`): | File | Covers | |---|---| | `ProxyTools` | Start/stop capture, proxy status, pause/resume | | `TrafficTools` | List/get requests, domains, traffic stats | | `FilterTools` | Filter requests, save/load/delete filter presets | | `AlertTools` | Manage alert rules, read alert events, toggle sound | | `AutoResponderTools` | Manage mock/auto-responder rules and status | | `ReplayTools` | Replay a captured request | | `SessionTools` | Save/load capture sessions | | `ExportTools` | Export as CSV / HAR / JSON / cURL / Postman | | `BookmarkTools` | Toggle/list bookmarks | | `ContentTools` | Encode/decode request or response bodies | | `StatsTools` | Traffic statistics, TLS info | Add it to an MCP-capable client (e.g. Claude Code / Claude Desktop) pointing at the built `HttpTrafficMonitor.McpServer.exe`; the main `HttpTrafficMonitor.exe` must already be running and capturing for most tools to return data. ### Configuring the MCP server Build it first: ``` dotnet build HttpTrafficMonitor.McpServer/HttpTrafficMonitor.McpServer.csproj -c Debug ``` This produces `HttpTrafficMonitor.McpServer/bin/Debug/net8.0/HttpTrafficMonitor.McpServer.exe`. Adjust paths below if you build in `Release` or from a different clone location. **Claude Code** — the repo ships a project-scoped `.mcp.json` in its root that starts the server with `dotnet run --project HttpTrafficMonitor.McpServer`, so opening the clone in Claude Code is enough (approve the `httptrafficmonitor` server when prompted) — no absolute paths to adjust. To register a built exe instead, use the CLI: ``` claude mcp add httptrafficmonitor -- "C:\Projects\httptrafficmonitr\HttpTrafficMonitor.McpServer\bin\Debug\net8.0\HttpTrafficMonitor.McpServer.exe" ``` which writes (or edit `.mcp.json` directly): ```json { "mcpServers": { "httptrafficmonitor": { "command": "C:\\Projects\\httptrafficmonitr\\HttpTrafficMonitor.McpServer\\bin\\Debug\\net8.0\\HttpTrafficMonitor.McpServer.exe", "args": [] } } } ``` **Claude Desktop** — add the same block to `claude_desktop_config.json` (`%APPDATA%\Claude\claude_desktop_config.json`): ```json { "mcpServers": { "httptrafficmonitor": { "command": "C:\\Projects\\httptrafficmonitr\\HttpTrafficMonitor.McpServer\\bin\\Debug\\net8.0\\HttpTrafficMonitor.McpServer.exe", "args": [] } } } ``` **Running via `dotnet` instead of the exe** (e.g. cross-platform host, or you'd rather not hardcode a self-contained exe path) — point `command` at `dotnet` and pass the DLL as the first arg: ```json { "mcpServers": { "httptrafficmonitor": { "command": "dotnet", "args": [ "C:\\Projects\\httptrafficmonitr\\HttpTrafficMonitor.McpServer\\bin\\Debug\\net8.0\\HttpTrafficMonitor.McpServer.dll" ] } } } ``` Optional environment variable (only needed if the main app's IPC API isn't on its default port): ```json { "mcpServers": { "httptrafficmonitor": { "command": "C:\\Projects\\httptrafficmonitr\\HttpTrafficMonitor.McpServer\\bin\\Debug\\net8.0\\HttpTrafficMonitor.McpServer.exe", "args": [], "env": { "HTM_API_URL": "http://localhost:18081" } } } } ``` After editing the config, restart Claude Code / Claude Desktop, launch `HttpTrafficMonitor.exe` (as Administrator) and click **Start**, then the `httptrafficmonitor` tools become available to the assistant. ## Requirements - Windows 10/11 - [.NET 8 SDK](https://dotnet.microsoft.com/download) (`net8.0-windows`, WPF) - **Administrator privileges** — the app requests elevation via its manifest (`app.manifest`, `requireAdministrator`), needed to register itself as the system proxy and install the local root CA > ⚠️ HTTPS interception installs a root CA into your system trust store and lets the app read HTTPS traffic in plaintext. Use it only for development/debugging on machines you control — the app shows a warning dialog on first launch and restores your original proxy settings when you stop monitoring or close it. ## Build & run ``` dotnet build HttpTrafficMonitor.csproj -c Debug ``` Run `bin/Debug/net8.0-windows/HttpTrafficMonitor.exe` (elevation prompt expected), or open `HttpTrafficMonitor.sln` in Visual Studio and hit F5. To build the MCP server: ``` dotnet build HttpTrafficMonitor.McpServer/HttpTrafficMonitor.McpServer.csproj ``` ## Tech stack - WPF on .NET 8, [WPF-UI](https://github.com/lepoco/wpfui) for Fluent Design - [Titanium.Web.Proxy](https://github.com/justcoding121/Titanium-Web-Proxy) for MITM proxying - [LiveChartsCore (SkiaSharp)](https://livecharts.dev/) for charts - [DiffPlex](https://github.com/mmanela/diffplex) for the Compare view - `ModelContextProtocol` .NET SDK for the MCP server ## Project structure ``` Models/ Request/session/alert/auto-responder/TLS data models Services/ Proxy, alerts, export, replay, session, content decoding, IPC API, theming ViewModels/ MVVM view models per view/panel Views/ MainWindow + dialogs (Advanced Filter, Replay/Compose, Comparison) HttpTrafficMonitor.McpServer/ Standalone MCP server (stdio) bridging to the running app's IPC API ``` ## UI automation Interactive controls across the main window and dialogs carry `AutomationProperties.AutomationId` (e.g. `Toolbar.StartButton`, `FilterBar.SearchTextBox`, `MainWindow.RequestsGrid`, `AdvancedFilter.ApplyFilterButton`, `AutoResponder.AddRuleButton`) so the app can be driven and tested via UI Automation / test tooling without relying on fragile name- or coordinate-based lookups.