Migrate to Vercel AI SDK and enable experimental multi-step tool execution#28
Merged
Kulikowski merged 15 commits intomainfrom Feb 25, 2026
Merged
Migrate to Vercel AI SDK and enable experimental multi-step tool execution#28Kulikowski merged 15 commits intomainfrom
Kulikowski merged 15 commits intomainfrom
Conversation
* De-dupe `findChromePath` logic * Replace custom backends with AI SDK
Kulikowski
changed the title
Contributor
Author
|
As discussed with @andreban we will add possibility of extending possible backends treating Vercel AI SDK as one of the possible backends, coming with next PRs :) |
andreban
approved these changes
Feb 25, 2026
Member
andreban
left a comment
andreban
left a comment
There was a problem hiding this comment.
LGTM, with comment on reintroducing the Backend abstraction in the next PR, so it's easy to add LLMs not supported by the Vercel SDK.
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
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Replaced direct API bindings with the AI SDK by Vercel, unlocking native structured tool-calling support.
Updated the evaluation loop and JSON format to support sequentially validating arrays of expectedCalls.
Modularized Evaluator: Split the monolithic
evaluator.tsinto a clean src/evaluator/ directory:models.ts: Instantiates LLM backends.
browser.ts: Extracts Puppeteer web integration.
mappers.ts: Normalizes raw extensions schemas into strict AI SDK interfaces.
prompts.ts: Extracts static system context.
index.ts: The core orchestrator.