Skip to content

feat: interactive browser session and supportsImages gating for browser_action #7198

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 5 commits into from
Closed

feat: interactive browser session and supportsImages gating for browser_action #7198

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
fe00413
feat: interactive browser session and supportsImages gating for brows…
hannesrudolph Aug 18, 2025
a87ef4a
chore: remove embedded tmp repos from index and ignore tmp/
hannesrudolph Aug 18, 2025
24d4e8c
chore: revert .gitignore change
hannesrudolph Aug 18, 2025
f67f466
Discard changes to .gitignore
hannesrudolph Aug 18, 2025
896a2e2
test: support remote browser connect/disconnect in BrowserSession and…
hannesrudolph Aug 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions .gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -46,3 +46,6 @@ logs
.qodo/
.vercel
.roo/mcp.json

# Ignore temporary clones
tmp/
2 changes: 1 addition & 1 deletion src/core/task/Task.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2207,7 +2207,7 @@ export class Task extends EventEmitter<TaskEvents> implements TaskLike {
return SYSTEM_PROMPT(
provider.context,
this.cwd,
(this.api.getModel().info.supportsComputerUse ?? false) && (browserToolEnabled ?? true),
(this.api.getModel().info.supportsImages ?? false) && (browserToolEnabled ?? true),
mcpHub,
this.diffStrategy,
browserViewportSize,
Expand Down
10 changes: 5 additions & 5 deletions src/core/webview/generateSystemPrompt.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,15 +45,15 @@ export const generateSystemPrompt = async (provider: ClineProvider, message: Web
const rooIgnoreInstructions = provider.getCurrentTask()?.rooIgnoreController?.getInstructions()

// Determine if browser tools can be used based on model support, mode, and user settings
let modelSupportsComputerUse = false
let modelSupportsImages = false

// Create a temporary API handler to check if the model supports computer use
// Create a temporary API handler to check if the model supports images (needed for screenshot-driven browsing)
// This avoids relying on an active Cline instance which might not exist during preview
try {
const tempApiHandler = buildApiHandler(apiConfiguration)
modelSupportsComputerUse = tempApiHandler.getModel().info.supportsComputerUse ?? false
modelSupportsImages = tempApiHandler.getModel().info.supportsImages ?? false
} catch (error) {
console.error("Error checking if model supports computer use:", error)
console.error("Error checking if model supports images:", error)
}

// Check if the current mode includes the browser tool group
Expand All @@ -62,7 +62,7 @@ export const generateSystemPrompt = async (provider: ClineProvider, message: Web

// Only enable browser tools if the model supports it, the mode includes browser tools,
// and browser tools are enabled in settings
const canUseBrowserTool = modelSupportsComputerUse && modeSupportsBrowser && (browserToolEnabled ?? true)
const canUseBrowserTool = modelSupportsImages && modeSupportsBrowser && (browserToolEnabled ?? true)

const systemPrompt = await SYSTEM_PROMPT(
provider.context,
Expand Down
Loading
Loading