The start page is the first-run setup flow users see before entering Home. It guides them through three steps: presentation, provider setup, and model configuration.
Step 1: Presentation
This step welcomes the user to Friday and explains the setup goal.
It is about:
- Introducing Friday.
- Explaining that setup connects an AI provider and chooses models.
- Letting the user start setup.
- Letting the user skip setup and go to Home.
Step 2: Provider Setup
This step lets the user connect at least one AI provider.
It is about:
- Showing available providers.
- Letting the user add or edit a provider API key.
- Showing whether a provider key is already saved.
- Keeping provider credentials in local app data.
- Continuing only after at least one provider is ready.
Step 3: Configure Models
This step lets the user review and configure the model areas documented in
docs/models. It should show the full model catalog while preserving each
module’s current support status.
It is about:
- Choosing the Friday Assistant large language model.
- Choosing the Voice Input speech-to-text model.
- Showing the Voice Output text-to-speech model and its runtime status.
- Showing image, video, and sound generation model areas with placeholder status until provider adapters are ready.
- Showing OCR endpoint/model status.
- Showing Embedding as unavailable until provider catalogs, vector index behavior, and runtime adapters are implemented.
The model areas covered by this step are:
| Model area | Purpose |
|---|---|
| Large language model | Main chat and agent reasoning model. |
| Speech to text | Model for dictation and transcription. |
| Text to speech | Model for spoken output. |
| Text to image | Model area for image generation. |
| Text to video | Model area for video generation. |
| Text to audio | Model area for sound and music generation. |
| OCR | Endpoint-backed document reading now, with future provider-backed model setup. |
| Embedding | Future semantic indexing model setup. |