Fix provider names to match OpenRouter API (z-ai, mistralai, moonshotai, deepseek, meta-llama, etc.)

Add OpenRouter support + 15+ popular models (Qwen, Grok, MiMo, Seed, Nemotron, GLM, Mercury, etc.)

Add support for multiple LLM providers (Zhipu, Moonshot, MiniMax, Mistral, Cohere, Groq, etc.)

Make LLM_MODEL required + add Duplicate Space button

Fix: use dangerouslyDisableDeviceAuth to bypass pairing for Docker

Remove invalid requireAuth config

Disable Control UI auth requirement for HF Spaces

Auto-inject gateway token into Control UI to bypass pairing

Pre-approve Control UI device in ~/.openclaw/devices/paired.json

Pre-create approved client entry to bypass pairing for Docker

Remove invalid pairing config - use default

Fix: use gateway.controlUi.pairing = false

Use dmPolicy: open to disable pairing for headless/Docker

Use --no-pairing CLI flag instead of config

Disable pairing for Docker/headless deployments

Fix: agents.defaults.model at top level, not under gateway

Set model via gateway.agents.defaults.model in config

Remove agents config - use environment variable for model instead

Add gateway logging and startup verification

Fix: agents should be object, not array

Remove LLM_PROVIDER entirely - auto-detect from model name

Fix: Support Gemini models properly with google/ prefix and correct agents config

Upload 12 files