ENยทESยทDEยทPTยทFR
โŒ˜K

First Run Wizard

On first launch (no config.toml found), aidaemon runs an interactive setup wizard.

Provider Selection

The wizard offers these presets:

ProviderBase URLDefault Models
Google AI Studio (Native)Native APIgemini-3-flash-preview / gemini-2.5-flash-lite / gemini-3-pro-preview
OpenAIhttps://api.openai.com/v1gpt-5-mini / gpt-5-nano / gpt-5.1
Anthropic (Native)Native APIclaude-sonnet-4-5 / claude-haiku-4-5 / claude-opus-4-6
Anthropic (OpenRouter)https://openrouter.ai/api/v1anthropic/claude-* variants
OpenRouterhttps://openrouter.ai/api/v1Mixed providers
Moonshot AI (Kimi)https://api.moonshot.ai/v1kimi-k2.5 / kimi-k2 / kimi-k2.5
MiniMaxhttps://api.minimax.io/v1MiniMax-M2.5 / MiniMax-M2.5-highspeed / MiniMax-M2.5
Cloudflare AI Gatewayhttps://gateway.ai.cloudflare.com/v1/<ACCOUNT_ID>/<GATEWAY_ID>/compatProvider-dependent (for example: gpt-5-mini / gpt-5-nano / gpt-5.1)
Ollama (local)http://localhost:11434/v1Auto-discovered from local instance
CustomUser-specifiedUser-specified

Wizard Steps

  1. Select provider โ€” choose from presets or enter custom
  2. Enter API key โ€” skipped for Ollama (no auth needed)
  3. Optional gateway token โ€” Cloudflare preset can add cf-aig-authorization for Authenticated Gateway mode
  4. Model selection โ€” auto-populated from preset, or auto-discovered for Ollama
  5. Telegram setup โ€” enter bot token and your numeric user ID
  6. Browser setup (if compiled with browser feature) โ€” auto-detects Chrome profiles
  7. Generate config.toml โ€” written to current directory
Ollama Auto-Discovery
When selecting Ollama, the wizard queries http://localhost:11434/api/tags to list all locally available models and lets you pick.

Chrome Profile Detection

The wizard auto-detects Chrome/Chromium profiles for browser session reuse:

  • macOS: ~/Library/Application Support/Google/Chrome
  • Linux: ~/.config/google-chrome or ~/.config/chromium

Reusing a profile inherits cookies and sessions, so the agent can access authenticated sites.