Canvas & Image Studio

The Canvas is Vurvey's chat workspace. It is the same core experience described on the Home page: a conversation surface where you work with agents, sources, tools, and image generation.

Image Studio is the image-editing experience that can open from chat-generated or uploaded images, and it also has a standalone page route in current vurvey-web-manager master.

Canvas

In current master:

• the Canvas loads the main chat view
• if chatbotEnabled is off for the workspace, the app redirects to Campaigns
• in guest mode, if there are no published personas, the page shows a No Brand Companions are currently available message

Current Chat Toolbar

The toolbar above the input uses a mix of labeled chips and icon-only controls.

Visible controls include:

Control	Current behavior
Agents	Opens the agent selector
Populations	Optional chip shown only when the related feature flag is enabled
Sources	Icon-only dropdown for attaching workspace sources
Images	Icon-only dropdown for image generation models
Tools	Icon-only dropdown for web and social research tools
Model selector	Optional icon-only selector shown only when the relevant feature flags are enabled

Sources

The Sources tooltip reads Select Datasets and/or Campaigns.

From the toolbar dropdown, the current quick actions are:

• Attach Datasets
• Attach Campaigns
• Turn on sources / Turn off sources

The full source modal supports campaign, dataset, and file/media flows. In the current UI, the top-level navigation is centered on Campaigns, Datasets, and All files, with question and dataset-media selection handled as drill-in pages rather than primary tabs.

Images

The image-model dropdown currently includes:

• Nano Banana
• OpenAI
• Google Imagen
• Stable Diffusion

The bottom option toggles image generation on or off.

Tools

The tools dropdown currently includes:

• Web Research
• TikTok
• Reddit
• LinkedIn
• YouTube
• X/Twitter
• Instagram

Older references to Google Trends, Google Maps, or Amazon as standard Home-toolbar tools are not accurate for the current master branch.

Other Chat Menus and Modals

The Canvas chat flow also uses a few other important interaction surfaces:

• the Agents chip opens the published-agent picker with filter chips, search, and explicit commit buttons
• the optional Populations chip opens the population-to-persona chooser with population search, persona drill-in, and Choose persona
• the upload button opens the shared Upload Files modal and then the Create new dataset modal
• typing / in the composer opens the contextual tool/image popup
• each assistant response includes a More dropdown and a delete confirmation dialog
• the conversation title/menu opens Rename conversation and Export conversation history to file dialogs

Image Studio

Image Studio is a real product surface in current master, but the supported actions are narrower and more concrete than some older docs suggested.

How It Opens

Current code supports:

• opening Image Studio from chat/image flows
• loading it as a standalone page through the image-studio route
• passing an image URL into the standalone page

Current Layout

The current layout has:

• a top navigation bar
• a left-side history rail
• the main editor area

Most editing controls are exposed as inline panels rather than separate popups. The main contextual surfaces are the history rail, the mask/selection controls, and the Convert to Video configuration panel.

Two other real popup/confirm surfaces in current master are:

• the brush-size popup inside the editor controls
• the unsaved-changes confirmation shown when you try to leave with edits

The top bar includes:

• Exit Image Studio
• Reset to Original when applicable
• Download
• an optional Save action when Image Studio is being used inside another flow

Editing Actions Verified in Master

The editor currently supports:

• selecting an area to change
• un-selecting parts of the mask
• resetting the current selection
• Enhance
• Upscale
• prompt-based image changes
• Remove on a masked area
• Convert to Video

Convert to Video

The current video conversion panel exposes these controls:

• Video prompt — text area describing the desired video content
• Duration — 4, 6, or 8 seconds
• Aspect ratio — 16:9 or 9:16
• Sample count — 1-4
• Person generation — Allow Adults or No People/Faces
• Negative prompt — text field for specifying what to exclude from the video
• Seed — numeric field for reproducible generation results
• Enhance prompt — checkbox to refine the prompt with Gemini before generation

History

Image Studio keeps a history rail of generated or edited media and can switch between image and video entries in that history.

What This Page No Longer Claims

To stay aligned with master, this guide intentionally does not document older speculative behavior such as:

• configurable Perlin-sphere workspace settings
• a guaranteed 9-card prompt showcase
• unsupported promises about five-panel studio layouts
• unverified feature lists beyond the actions currently wired in the editor