-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Update CerebrasClient to use ai-sdk instead of OpenAI #1148
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.
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
base: main
Are you sure you want to change the base?
Conversation
# why solves browserbase#1060 patch regression of playwright arguments being removed from agent execute response # what changed agent.execute now returns playwright arguments in its response # test plan tested locally
…ms to docs (browserbase#1065) # why reflect project id changes in docs # what changed advanced configuration comments # test plan reviewed via mintlify on localhost
# why Easier to use for Custom LLM Clients and keep users up to date with our aisdk file # what changed added export of aisdk to lib/index.ts # test plan build local stagehand, import local AISdkClient, run Azure Stagehand session
…onfigu… (browserbase#1073) …ration settings # why Updated docs to match the new fingerprint params in the Browserbase docs here: https://docs.browserbase.com/guides/stealth-customization#customization-options # what changed Update browser configuration docs to reflect the docs changes. # test plan
# why Updating docs to reflect aisdk can be imported directly # what changed The model page # test plan Reviewed page with mintlify dev locally
# why # what changed # test plan
# why Currently, we do not support stagehand agent within the api # what changed When api is enabled, stagehand agent now routes through the api # test plan Tested locally
# why Currently, using playwright screenshot command is not available when the execution environment is Stagehand. A customer has indicated they would prefer to use Playwright's native screenshot command instead of CDP when using Browserbase as CDP screenshot causes unexpected behavior for their target site. # what changed - added a StagehandScreenshotOptions type with useCDP argument added - extended page type to accept custom stagehand screeenshot options - update screenshot proxy to default useCDP to true if the env is browserbase and use playwright screenshot if false - added eval for screenshot with and without cdp # test plan - tested and confirmed functionality with eval and external example script (not committed)
…rowserbase#1057) # why We want to build a best in class agent in stagehand. Therefore, we need more eval benchmarks. # what changed - Added Web-bench evals dataset - Added a subset of OS World evals - those that can be run in a chrome browser (desktop-based tasks omitted) - added LICENSE noticed to the copied evals tasks - Added ground truth / expected result to some WebVoyager tasks using reference_answer.json from Browser Use public evals repo. Improvements to `pnpm run evals -man` to better describe how to run evals. # test plan Evals should run locally and bb for these new benchmarks.
# why Initial instructions didn't mention uv or pip prerequisites and also didn't mention venv. Fix reduces friction on first timers. # what changed - added link to install uv - added details for initializing venv - adjusted code example respectively # test plan docs change
# why - webpage structure changed, needed to update the xpath in the expected locator
… with LanguageModelV1 + LiteLLM works for python (browserbase#1086) # why 1. aisdk not yet available through npm package 2. customLLM provider only works with LanguageModelV1 3. LiteLLM compatible providers are supported in python # what changed 1. change docs to install stagehand from git repo 2. pin versions that use LanguageModelV1 # test plan local test
# why currently we pass stagehand page to agent, this results in our page management having issues when facing new tabs # what changed the stagehand object is now passed instead of stagehandPage # test plan tested locally
# why Our existing screenshot service is a dummy time-based triggered service. It also does not trigger based on any actions of the agent. # what changed Added img hash diff algo (quick check with MSE, verify with SSIM algo) to see if there was an actual UI change and only store ss in the buffer if that is so. Added ss interceptor which copies each screenshot the agent is taking to a buffer (if different enough from the previous ss) to be later used for evals. - There's also a small refactor of the agent initialization config to enable the screenshot collector service to be attached # test plan Tests pass locally --------- Co-authored-by: Miguel <36487034+miguelg719@users.noreply.github.com> Co-authored-by: miguel <miguelg71921@gmail.com>
# why To help make sense of eval test cases and results # what changed Added metadata to eval runs, cleaned deprecated code # test plan
# why # what changed # test plan
# why anthropic released a new sota computer use model # what changed added claude-sonnet-4-5-20250929 as a model to the list # test plan ran evals
…ase#1103) Why Custom AI SDK tools and MCP integrations weren't working properly with Anthropic CUA - parameters were empty {} and tools weren't tracked. What Changed - Convert Zod schemas to JSON Schema before sending to Anthropic (using zodToJsonSchema) - Track custom tool calls in the actions array - Silence "Unknown tool name" warnings for custom tools Test Plan Tested with examples file. Parameters passed correctly ({"city":"San Francisco"} instead of {}) Custom tools execute and appear in actions array No warnings
# why To improve context # what changed Added current page and url to the system prompt # test plan
# why To inform the user throughout the agent execution process # what changed Added logs to tool calls, and on the stagehand agent handler # test plan - [x] tested locally
PR to make clearer the dependencies for `extract` (for those who haven't used zod or pydantic before) --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
# why
- before this change, when we convert `z.string().url()` to an ID, if it
was inside a `z.array()`, it was not getting converted back into a URL
- this meant that if you defined a schema like this:
```ts
schema: z.object({
records: z.array(z.string().url()),
})
```
you would receive an array like this:
```
{
records: [
'0-302', '0-309',
'0-316', '0-323',
'0-330', '0-337',
'0-344', '0-351',
'0-358', '0-365'
]
}
```
- with this change, you will now receive the actual URLs, ie:
```
{
records: [
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10003-10041.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10004-10143%20(C06932208).pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10004-10143.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10004-10156.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10004-10213.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10005-10321.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10006-10247.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10007-10345.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10009-10021.pdf',
'https://www.archives.gov/files/research/jfk/releases/2025/0318/104-10009-10222.pdf'
]
}
```
# what changed
- updated the `injectUrls` function so that when it hits an array and
there is not deeper path, it loops through the array and injects the
URLs
# test plan
- evals
# why Adding support for Gemini's new Computer Use model # what changed We partnered with Google Deepmind to help integrate and test their new Computer Use models. <img width="1238" height="655" alt="Screenshot 2025-10-07 at 1 14 44 PM" src="https://github.com/user-attachments/assets/af0d854a-8e55-4937-a071-10335497f686" /> The new model tag `gemini-2.5-pro-computer-use-preview-10-2025` is available for Stagehand Agent. You can try it today with the example `cua-example.ts` To learn more, check out the blog post [https://www.browserbase.com/blog/evaluating-browser-agents](https://www.browserbase.com/blog/evaluating-browser-agents) --------- Co-authored-by: tkattkat <tkat@tkat.net> Co-authored-by: Kylejeong2 <kylejeong21@gmail.com> Co-authored-by: Sameel <sameel.m.arif@gmail.com>
# why # what changed # test plan
This PR was opened by the [Changesets release](https://github.com/changesets/action) GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated. # Releases ## @browserbasehq/stagehand@2.5.1 ### Patch Changes - [browserbase#1082](browserbase#1082) [`8c0fd01`](browserbase@8c0fd01) Thanks [@tkattkat](https://github.com/tkattkat)! - Pass stagehand object to agent instead of stagehand page - [browserbase#1104](browserbase#1104) [`a1ad06c`](browserbase@a1ad06c) Thanks [@miguelg719](https://github.com/miguelg719)! - Fix logging for stagehand agent - [browserbase#1066](browserbase#1066) [`9daa584`](browserbase@9daa584) Thanks [@tkattkat](https://github.com/tkattkat)! - Add playwright arguments to agent execute response - [browserbase#1077](browserbase#1077) [`7f38b3a`](browserbase@7f38b3a) Thanks [@tkattkat](https://github.com/tkattkat)! - adds support for stagehand agent in the api - [browserbase#1032](browserbase#1032) [`bf2d0e7`](browserbase@bf2d0e7) Thanks [@miguelg719](https://github.com/miguelg719)! - Fix for zod peer dependency support - [browserbase#1014](browserbase#1014) [`6966201`](browserbase@6966201) Thanks [@tkattkat](https://github.com/tkattkat)! - Replace operator handler with base of new agent - [browserbase#1089](browserbase#1089) [`536f366`](browserbase@536f366) Thanks [@miguelg719](https://github.com/miguelg719)! - Fixed info logs on api session create - [browserbase#1103](browserbase#1103) [`889cb6c`](browserbase@889cb6c) Thanks [@tkattkat](https://github.com/tkattkat)! - patch custom tool support in anthropic cua client - [browserbase#1056](browserbase#1056) [`6a002b2`](browserbase@6a002b2) Thanks [@chrisreadsf](https://github.com/chrisreadsf)! - remove need for duplicate project id if already passed to Stagehand - [browserbase#1090](browserbase#1090) [`8ff5c5a`](browserbase@8ff5c5a) Thanks [@miguelg719](https://github.com/miguelg719)! - Improve failed act error logs - [browserbase#1014](browserbase#1014) [`6966201`](browserbase@6966201) Thanks [@tkattkat](https://github.com/tkattkat)! - replace operator agent with scaffold for new stagehand agent - [browserbase#1107](browserbase#1107) [`3ccf335`](browserbase@3ccf335) Thanks [@seanmcguire12](https://github.com/seanmcguire12)! - fix: url extraction not working inside an array - [browserbase#1102](browserbase#1102) [`a99aa48`](browserbase@a99aa48) Thanks [@miguelg719](https://github.com/miguelg719)! - Add current page and date context to agent - [browserbase#1110](browserbase#1110) [`dda52f1`](browserbase@dda52f1) Thanks [@miguelg719](https://github.com/miguelg719)! - Add support for new Gemini Computer Use models ## @browserbasehq/stagehand-evals@1.1.0 ### Minor Changes - [browserbase#1057](browserbase#1057) [`b7be89e`](browserbase@b7be89e) Thanks [@filip-michalsky](https://github.com/filip-michalsky)! - added web voyager ground truth (optional), added web bench, and subset of OSWorld evals which run on a browser ### Patch Changes - [browserbase#1072](browserbase#1072) [`dc2d420`](browserbase@dc2d420) Thanks [@filip-michalsky](https://github.com/filip-michalsky)! - improve evals screenshot service - add img hashing diff to add screenshots and change to screenshot intercepts from the agent - Updated dependencies \[[`8c0fd01`](browserbase@8c0fd01), [`a1ad06c`](browserbase@a1ad06c), [`9daa584`](browserbase@9daa584), [`7f38b3a`](browserbase@7f38b3a), [`bf2d0e7`](browserbase@bf2d0e7), [`6966201`](browserbase@6966201), [`536f366`](browserbase@536f366), [`889cb6c`](browserbase@889cb6c), [`6a002b2`](browserbase@6a002b2), [`8ff5c5a`](browserbase@8ff5c5a), [`6966201`](browserbase@6966201), [`3ccf335`](browserbase@3ccf335), [`a99aa48`](browserbase@a99aa48), [`dda52f1`](browserbase@dda52f1)]: - @browserbasehq/stagehand@2.5.1 ## @browserbasehq/stagehand-examples@1.0.10 ### Patch Changes - Updated dependencies \[[`8c0fd01`](browserbase@8c0fd01), [`a1ad06c`](browserbase@a1ad06c), [`9daa584`](browserbase@9daa584), [`7f38b3a`](browserbase@7f38b3a), [`bf2d0e7`](browserbase@bf2d0e7), [`6966201`](browserbase@6966201), [`536f366`](browserbase@536f366), [`889cb6c`](browserbase@889cb6c), [`6a002b2`](browserbase@6a002b2), [`8ff5c5a`](browserbase@8ff5c5a), [`6966201`](browserbase@6966201), [`3ccf335`](browserbase@3ccf335), [`a99aa48`](browserbase@a99aa48), [`dda52f1`](browserbase@dda52f1)]: - @browserbasehq/stagehand@2.5.1 Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
# why The original example used JavaScript destructuring syntax [table] which doesn't work in Python. Fixed to use proper Python array indexing. # what changed fixed example to proper python syntax # test plan Co-authored-by: Steven Bryan <steven@mac.local.meter>
# why - need to set default viewport when running on browserbase. previously, we only defined the default inside the exported `StagehandConfig` # what changed - set default viewport to 1288 * 711 when running on browserbase # test plan - tested locally, - regression evals
This PR was opened by the [Changesets release](https://github.com/changesets/action) GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated. # Releases ## @browserbasehq/stagehand@2.5.2 ### Patch Changes - [browserbase#1114](browserbase#1114) [`c0fbc51`](browserbase@c0fbc51) Thanks [@seanmcguire12](https://github.com/seanmcguire12)! - configure default viewport when running on browserbase ## @browserbasehq/stagehand-evals@1.1.1 ### Patch Changes - Updated dependencies \[[`c0fbc51`](browserbase@c0fbc51)]: - @browserbasehq/stagehand@2.5.2 ## @browserbasehq/stagehand-examples@1.0.11 ### Patch Changes - Updated dependencies \[[`c0fbc51`](browserbase@c0fbc51)]: - @browserbasehq/stagehand@2.5.2 Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Updated link in the Getting Started section to point to the correct Quickstart Guide. # why Quickstart link in README leads to a non-existent page. <img width="1556" height="763" alt="image" src="https://github.com/user-attachments/assets/20a1a5b5-8534-43b4-89d5-e3a062b3965a" /> # what changed Updated quickstart link in README to the correct quickstart address `https://docs.stagehand.dev/first-steps/quickstart` # test plan Access new link to quickstart
# why currently, for openai cua agent, we are handling keypress actions incorrectly currently, there is no way to pass a custom system prompt to the Google cua agent # what changed - All key actions, are now ran through mapKeyToPlaywright function to ensure we are properly mapping the agents actions to valid playwright keys - Custom system prompts now override the default system prompt for Google Cua agent # test plan tested locally with google & openai cua agents Fixes browserbase#1122
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Greptile Overview
Summary
Migrated Cerebras integration from OpenAI client to AI SDK's official Cerebras provider, adding support for 7 new Cerebras models.
Major changes:
- Replaced custom OpenAI-based implementation in
lib/llm/CerebrasClient.tswith AI SDK wrapper usingcreateCerebrasprovider - Simplified from ~310 lines to ~70 lines by delegating to
AISdkClient - Added new models:
gpt-oss-120b,llama-4-maverick-17b-128e-instruct,llama-4-scout-17b-16e-instruct,qwen-3-235b-a22b-instruct-2507,qwen-3-235b-a22b-thinking-2507,qwen-3-32b,qwen-3-coder-480b - Exported
CerebrasClientin public API (lib/index.ts)
Issues found:
- Critical bug: Logger implementation references non-existent
this.loggerproperty, will cause runtime error - Config file (
stagehand.config.ts) mistakenly changed to test Cerebras model instead of production default
Confidence Score: 1/5
- This PR has critical bugs that will cause runtime errors and should not be merged without fixes
- Score reflects a critical logic error in CerebrasClient.ts:41-50 where the logger implementation references a non-existent
this.loggerproperty. This will cause a runtime error when the logger is invoked. Additionally, the config file changes should be reverted. Missing changeset also violates repository requirements. - Pay close attention to
lib/llm/CerebrasClient.ts(logger bug) and revertstagehand.config.tsto production defaults
Important Files Changed
File Analysis
| Filename | Score | Overview |
|---|---|---|
| lib/llm/CerebrasClient.ts | 2/5 | Refactored to use AI SDK client with Cerebras provider instead of OpenAI client, but contains a critical bug in logger implementation |
| lib/llm/LLMProvider.ts | 4/5 | Added type casts for Cerebras provider and added model mappings for 7 new Cerebras models |
| stagehand.config.ts | 5/5 | Updated default config to use Cerebras model for testing purposes |
Sequence Diagram
sequenceDiagram
participant User
participant LLMProvider
participant CerebrasClient
participant AISdkClient
participant CerebrasProvider
participant CerebrasAPI
User->>LLMProvider: getClient("cerebras-qwen-3-32b")
LLMProvider->>LLMProvider: Check modelToProviderMap
LLMProvider->>CerebrasClient: new CerebrasClient(options)
CerebrasClient->>CerebrasClient: Extract model name (remove "cerebras-" prefix)
CerebrasClient->>CerebrasProvider: createCerebras({apiKey})
CerebrasProvider-->>CerebrasClient: provider instance
CerebrasClient->>CerebrasProvider: provider(modelName)
CerebrasProvider-->>CerebrasClient: LanguageModel
CerebrasClient->>AISdkClient: new AISdkClient({model, logger, cache})
AISdkClient-->>CerebrasClient: client instance
CerebrasClient-->>LLMProvider: CerebrasClient instance
LLMProvider-->>User: LLMClient
User->>CerebrasClient: createChatCompletion(options)
CerebrasClient->>AISdkClient: createChatCompletion(options)
AISdkClient->>AISdkClient: Format messages to CoreMessage[]
AISdkClient->>CerebrasAPI: generateObject/generateText
CerebrasAPI-->>AISdkClient: response
AISdkClient->>AISdkClient: Transform to LLMResponse format
AISdkClient-->>CerebrasClient: LLMResponse
CerebrasClient-->>User: LLMResponse
Additional Comments (2)
-
lib/llm/CerebrasClient.ts, line 41-50 (link)logic: The logger passed to
AISdkClientconstructor is immediately invoked inline, butthis.loggerdoesn't exist yet (not defined in the class). This code transforms log messages but the original logger isn't stored anywhere to call. -
stagehand.config.ts, line 10-12 (link)logic: Config file should be reverted to use default production settings, not test configuration from this PR
5 files reviewed, 2 comments
|
test |
why
Cerebras integration was out of date for models and not buildable. AI-SDK ships a first‑party Cerebras provider that's now used.
what changed
test plan
I tested using the form_filling_sensible file and the 2048 tests to test all the Cerebras models.
notes
Message & tool mapping mirrors the existing AISdkClient example for consistent behavior. No changes to other providers.