A transparent proxy system for inspecting, debugging, and modifying any traffic from the agent container, focused on LLM-based coding agents. Works with Claude Code, Codex CLI, Cursor, Aider, and any other tool that calls LLM APIs.
# Tollbooth A transparent proxy for inspecting, debugging, and modifying network traffic—built for LLM agent research but applicable to any HTTP/HTTPS traffic. Works with **Claude Code**, **Codex CLI**, **Cursor**, **Aider**, and any tool that makes network requests. > **[Read the full documentation](https://flechettelabs.com/tollbooth)** ## What It Does Tollbooth is a man-in-the-middle proxy that captures **all network traffic** from the agent container. While built for monitoring LLM agent communications (with automatic parsing of API calls into conversations), it intercepts everything—including requests from tools that agents invoke, package managers, and any other network activity. Useful for agent research, security testing, and network debugging. - **Inspect All Traffic** - See every HTTP request. LLM API calls are parsed into readable conversations; other traffic shown raw. - **Conversation Trees** - Visualize branching conversations with a gitflow-style tree view. Branches from retries, replays, and natural divergence are shown with merge connectors where paths reconverge. - **Modify Requests** - Intercept and edit any request before it reaches its destination. Mark items as timeout-immune for extended inspection. - **Mock Responses** - Serve stored responses instead of forwarding requests - **Rules Engine** - Automated traffic handling with priority-based rules, static modifications, drop actions, and smart fall-through when modifications don't apply - **Export Conversations** - Export conversations as JSON, Markdown, or styled HTML - **Detect Refusals** - ML-powered detection of LLM refusals with automatic handling ## Quick Start ```bash # 1. Generate certificates ./setup-certs.sh # 2. Start services docker compose up # 3. Open UI open http://localhost:5173 # 4. Run your agent through the proxy docker compose run --rm agent ``` ## Architecture ``` ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐ │ LLM Agent │────▶│
Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.