Skip to content

feat(browser-runtime): add Browser Use as a managed runtime - #395

Draft
gregpr07 wants to merge 1 commit into
TIGER-AI-Lab:mainfrom
gregpr07:add-browser-use
Draft

gregpr07 wants to merge 1 commit into
TIGER-AI-Lab:mainfrom
gregpr07:add-browser-use

Conversation

@gregpr07

Copy link
Copy Markdown

What does this PR do?

Hey! Kernel and Browserbase are already runtimes here, so this adds Browser Use Cloud the same way, following the Kernel provider: --browser-runtime browser-use creates a browser via POST /api/v4/browsers (BROWSER_USE_API_KEY), resolves its CDP WebSocket, stops it after the run, and downloads the recording to data/recording.mp4. Wired into run, batch, and the TUI like the other managed runtimes (concurrency default 1, no chrome-extension harness), plus docs, changelog, and mocked unit tests. No new dependencies.

Corpus

  • v2
  • v1
  • both
  • not applicable

Test plan

  • uv run --frozen ruff check . / ruff format --check . / pyright src/clawbench tests: clean
  • uv run --frozen pytest: 337 passed, 3 skipped (new tests mock all HTTP, no live API calls)
  • scripts/ci/check_markdown_links.py: all links resolve
  • Not yet run against a live Browser Use browser; happy to do a real batch before review if you want

Related issues

Part of #103

🤖 Generated with Claude Code

Adds `--browser-runtime browser-use`, following the Kernel provider:
create a Browser Use Cloud browser over the v4 REST API, resolve its
CDP WebSocket, stop it after the run, and download the session
recording to data/recording.mp4. Wired into run, batch, and the TUI
like the other managed runtimes, with docs and mocked unit tests.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant