Add Meta Computer Use agent - #116
Conversation
Expose start, status, get, stop, and startAndWait for Muse Spark 1.1 computer-use tasks. Co-authored-by: nikhil <nikhil@metlo.com>
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
There was a problem hiding this comment.
NikhilShahi has reached the 50-credit limit for trial accounts. To continue receiving code reviews, upgrade your plan.
Co-authored-by: nikhil <nikhil@metlo.com>
Align Meta Computer Use LLM and reasoning-effort types with the current API. Co-authored-by: nikhil <nikhil@metlo.com>
There was a problem hiding this comment.
NikhilShahi has reached the 50-credit limit for trial accounts. To continue receiving code reviews, upgrade your plan.
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.
❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.
Want higher recall? High effort reviews run extra passes and find more bugs. A team admin can switch effort levels in the Cursor dashboard.
Reviewed by Cursor Bugbot for commit 5283c91. Configure here.
|
|
||
| export type MetaComputerUseLlm = "muse-spark-1.1" | "muse-spark-1.2" | "muse-spark-1.3"; | ||
|
|
||
| export type MetaReasoningEffort = "minimal" | "low" | "medium" | "high" | "xhigh" | "max"; |
There was a problem hiding this comment.
Invalid reasoning effort value allowed
Medium Severity
MetaReasoningEffort includes max, which is not a valid Muse Spark reasoning effort. The supported values are minimal, low, medium, high, and xhigh. Callers who pick max from the public type will send a value the API rejects.
Additional Locations (1)
Reviewed by Cursor Bugbot for commit 5283c91. Configure here.
Timestamps are unix integers, errors are objects, and reasoning uses effort/summary instead of the Grok string/any placeholders. Co-authored-by: nikhil <nikhil@metlo.com>
There was a problem hiding this comment.
NikhilShahi has reached the 50-credit limit for trial accounts. To continue receiving code reviews, upgrade your plan.
The production get payload includes createdAt, finishedAt, liveDomain, and jobParams in addition to the compacted step history. Co-authored-by: nikhil <nikhil@metlo.com>
There was a problem hiding this comment.
NikhilShahi has reached the 50-credit limit for trial accounts. To continue receiving code reviews, upgrade your plan.
Keep the task response aligned with Grok/CUA: no createdAt, finishedAt, liveDomain, or jobParams. Co-authored-by: nikhil <nikhil@metlo.com>
There was a problem hiding this comment.
NikhilShahi has reached the 50-credit limit for trial accounts. To continue receiving code reviews, upgrade your plan.
2 participants
Summary
Adds Meta Computer Use to the Node SDK, matching Grok Computer Use and the current Hyperbrowser API (
muse-spark-1.1/1.2/1.3, includingmaxreasoning on 1.3).Changes
client.agents.metaComputerUsewithstart,get,getStatus,stop, andstartAndWaituseCustomApiKeys+apiKeys.metataskis required{code,message}errors, and{effort,summary}reasoningjobId,status,metadata,data,error,liveUrl)Validation
yarn lintyarn buildyarn test:integrationstart,getStatus,get,stop, andstartAndWaitagainsthttps://api.hyperbrowser.ai