LLM API Key Proxy (Fork)

A personal fork of Mirrowel/LLM-API-Key-Proxy with additional providers, fixes, and tooling.

For full documentation, see the upstream repository.

---

Fork-Specific Features

Additional Providers

Provider	Description
Cursor	Cursor IDE API with quota monitoring integration
ZenMux	OpenAI-compatible provider with custom header support for free models
DedalusLabs	OpenAI-compatible provider with tool_choice handling

Fixes & Improvements

• Streaming token counting — Correct input_tokens in message_start for Claude Code statusline
• Anthropic null response handling — Defensive checks for empty/invalid responses
• Symlink atomic writes — Resolve symlinks before writes for Docker volume mounts
• Dynamic provider singleton fix — Prevents api_base sharing between dynamic providers
• Kimi K2.5 transform — Auto-enforces top_p=0.95 for Kimi models
• Quota group sync — Dynamic model discovery for accurate quota tracking

Tooling

• Transaction Log Viewer TUI — Browse and inspect API request/response logs
• Cursor Sidecar — OpenAI-compatible proxy for Cursor API
• Embedding Support — Dispatch embeddings to appropriate providers

---

Quick Start (Docker)

docker-compose up -d

Or use the Komodo stack for deployment.

Environment Variables

See upstream documentation for base configuration. Fork-specific variables:

# Cursor provider
CURSOR_API_KEY_1=your-cursor-key

# ZenMux (free models)
ZENMUX_API_BASE=https://zenmux.example.com/v1
ZENMUX_API_KEY_1=your-zenmux-key

# DedalusLabs
DEDALUSLABS_API_BASE=https://api.dedaluslabs.ai/v1
DEDALUSLABS_API_KEY_1=dsk-live-xxxxx

---

Upstream Sync

This fork is regularly synced with upstream. See .agent/skills/upstream-sync/ for the sync workflow.

---

License

Same as upstream — see LICENSE.