Settings
Tune the assistant under the hood.
Configure brains, API keys, PLM Network, Google Workspace, and Sqwibbl’s editable identity without exposing that complexity in the everyday launcher.
Configuration cards
Choose what to set up.
Sqwibbl brain/model
Default model choice used by the launcher beside the Ask Sqwibbl button.
Local Brain
Model path, GGUF file, chat template, context size, and GPU layers.
checking…Open setup →Anthropic / Claude
API key, model selection, and token limits for Claude frontier access.
checking…Open setup →PLM Brain
PLM-hosted OpenAI-compatible brain endpoint, model path, and brain ID.
checking…Open setup →OpenAI
OpenAI API key, model, and endpoint for GPT brain access.
checking…Open setup →MiniMax M2.7
Built-in large LLM fallback via MiniMax's OpenAI-compatible chat endpoint.
checking…Open setup →Google Gemini
Gemini API key, model, and OpenAI-compatible endpoint for the launcher brain picker.
checking…Open setup →Grok xAI
xAI API key, Grok model, and endpoint for the launcher brain picker.
checking…Open setup →HuggingFace
Hub model discovery, Serverless Inference, and approved Inference Providers for specialist tasks.
checking…Open setup →PLM Network
Register with PLMMarket, set your JWT token, and configure agent identity.
checking…Open setup →Add Funds
Fund this Sqwibbl’s PLMMarket Circle wallet through Stripe crypto onramp.
Wallet fundingOpen funding →Google Workspace
Connect Gmail, Drive, Sheets, and Docs with local OAuth credentials.
checking…Open setup →Social Media
Connect TikTok, Facebook, X, LinkedIn, and YouTube accounts through Composio OAuth.
checking…Open setup →SOUL.md Editor
Edit the assistant persona, heartbeat interval, and brain routing live.
Live editorOpen editor →