ccglass

ccglass

See exactly what your coding agent sends to the model. A lightweight
local logging reverse-proxy + web dashboard for Claude Code, Codex,
OpenCode, DeepSeek-TUI, Reasonix, CodeBuddy, Kimi, Ollama, OpenRouter, and more.
One command, like ollama:

Run with no arguments and ccglass asks which client to inspect:

Or name it directly: ccglass claude, ccglass codex, ccglass deepseek,
ccglass deepseek-tui, ccglass reasonix, ccglass dsnix, or ccglass kimi.

ccglass starts a proxy, points the client at it via the right base-URL env var,
launches it for you, and opens a dashboard where you watch every request in real
time — the full system prompt, every tool schema, the message history,
token/cache/cost numbers, and a turn-to-turn diff.

Why

These CLIs are Node/native apps that ignore HTTP_PROXY/HTTPS_PROXY — so
Charles/mitmproxy never see the traffic, and fetch-patching tools break across
updates. ccglass sidesteps all of it: the client does the HTTPS to the real API
itself; you only intercept the plain HTTP hop to localhost. No CA certs, no TLS
pinning.

Providers

Notes by provider:

- Codex — captures traffic when Codex is in API-key mode (OPENAI_API_KEY). If Codex is authenticated via ChatGPT login, it uses a WebSocket transport (wss://chatgpt.com/...) that bypasses OPENAI_BASE_URL — the dashboard will be empty. Run codex doctor to check your auth mode; if it shows auth mode: chatgpt, switch to API-key mode to use ccglass.
- Kimi — runs through Claude Code against Moonshot's Anthropic-compatible endpoint; set ANTHROPIC_AUTH_TOKEN to your Moonshot key.
- DeepSeek-TUI — OpenAI-compatible Chat Completions; set DEEPSEEK_API_KEY.
- Reasonix — OpenAI-compatible Chat Completions; set DEEPSEEK_API_KEY.
- OpenCode — auto-detects upstream from OPENAI_BASE_URL; set it before…