Add Parallel as a search option - #38
Open
georgeatparallel wants to merge 1 commit into
Open
georgeatparallel wants to merge 1 commit into
georgeatparallel wants to merge 1 commit into
Conversation
georgeatparallel
marked this pull request as ready for review
September 22, 2026 00:22
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Open Browser already has a
searchaction, but its browser search engines return a results page that the agent has to read. This addsengine: "parallel"to that action. It returns source URLs and excerpts directly through the command result, without a Parallel account or API key. Google remains the default, and the existing engines andweb_searchaction keep their behavior.The anonymous Parallel Search MCP uses Fast. Artificial Analysis's DeepSearchQA results report 80 F1 for Parallel Fast, and its Time per Task chart includes both model and search time across benchmark tasks. Those are provider benchmarks, not measurements of Open Browser. They make a useful case for offering keyless search alongside browser navigation.
The new route uses the MCP SDK's Streamable HTTP client, calls
web_search, and puts up to five cited results into the existing action result. It checks the configured URL restrictions before connecting. Queries and an objective based on each query go to Parallel; browser cookies and page content don't. Requests identify the project with anopen-browser/<version>User-Agent for aggregate usage measurement. The README explains the choice and the data flow.Validation: the new action tests pass (4/4), the core suite passes (368/368), and all three packages build. A live call through
CommandExecutor.executeActionreturned five cited results from the anonymous endpoint. Biome passes on the new adapter and its tests. The repo-wide test script exits 1 because the CLI and sandbox packages have no test files; repo-wide lint reports 180 existing errors in unchanged files.Btw - I'm a devrel engineer at Parallel