Repository navigation
feat(integrations): add Grok Build CLI harness - #3120
antonvishal wants to merge 23 commits into
Conversation
Resolve overview.mdx conflicts to include both Grok Build and fx entries. Keep the screenshot base64 budget from browserbase#2791 alongside the ACP facade launcher changes. Also pass --max-screenshot-base64-bytes=60000 to the facade spawned by the ACP launcher so ACP transports get the same frame-cap protection as fx, mark BROWSERBASE_PROJECT_ID optional in the Grok Build README per browserbase#2777, and update the affected ACP arg assertions.
# Conflicts: # packages/docs/docs.json # packages/docs/v4/integrations/overview.mdx
…2-pr-2876 # Conflicts: # turbo.json
Bring PR browserbase#2876 onto Miguel's facade/eval stack: native `--rules` policy, turn budget, session-loss, keyed observations, and reported usage/cost from the grok CLI stream. Co-authored-by: Cursor <cursoragent@cursor.com>
The short dependency-wait timeout can abort after fetch has already been sent. That late hit was counted in the next phase and failed TypeScript unit under CI load. Co-authored-by: Cursor <cursoragent@cursor.com>
…sertion" This reverts commit 215a91a.
# Conflicts: # packages/evals/package.json # pnpm-lock.yaml
|
|
This PR is from an external contributor and must be approved by a stagehand team member with write access before CI can run. |
github-actions
Bot
added
external-contributor
This branch has not been deployed
1 participant
Why
Let evals run through the native Grok Build CLI and share Stagehand browser evidence. Replaces #2876 after review requested that workflow changes move to the mirror PR.
What
Adds the Grok Build session adapter, eval harness, configuration example, and docs. Supports configurable auto approval and preserves tool results, usage, cost, and stop reasons.
How
Mounts the shared browser runtime through trusted MCP configuration in an isolated Grok home. No .github changes remain in the PR diff.
Summary by cubic
Adds a Grok Build CLI integration so evals can run through the native
grokbinary and share Stagehand browser evidence, replacing PR #2876.New Features
grok-buildeval harness, the@browserbasehq/stagehand-integrations-grok-build-sdksession adapter, config example, and docs.--always-approve, tool results, usage, cost, and stop reasons.grok_buildas a concrete executable harness withstagehand_facade,playwright_mcp, andchrome_devtools_mcptool surfaces.Written for commit b09c74b. Summary will update on new commits.