Adding an HTTP Proxy to Coding Agents
dandriscoll · hn · 2026-07-14
The author developed Trollbridge, a front-end control layer for coding agents, preventing them from having completely unrestricted access to external HTTP/S resources.
Core Features
- Allow/deny lists
- Real-time request approval or blocking within the TUI
- Optional LLM integration to assist with approvals based on your past decisions
Background
The author originally used an Incus VM for local sandboxing but found that agents constantly required new permissions, creating security holes. Additionally, traditional firewalls don't distinguish between HTTP methods. Trollbridge aims to let you "guide" the LLM rather than granting it full autonomy.
More from coding & agent
- OpenAI ships GPT-6.1 Terra; Brockman says its computer use now beats his own speed — gdb · 2026-10-27
- Kolab Benchmark Uses Pro Screen Recordings to Show AI Agents Overestimate Collaboration Quality — minnesotanlp · 2026-10-27
- Dev argues Claude burns far more tokens than GPT models for the same coding work — haider1 · 2026-10-27
- Stop building agents that can't change models: stable contracts, external state, real evals — rseroter · 2026-10-27
- Cheap models can cost more in agent harnesses than frontier ones, researcher warns — vasuman · 2026-10-27
- Kanzo: A 5k-Line Rust Kanban Binary With MCP for Claude Code and Codex — allmudi · 2026-10-27