Pull Request Explorer

Summary

Health Assessment

Summary

Accept marketplace integration terms with a single confirmation

Simplifies the CLI integration acceptance flow by consolidating multiple legal document prompts into a single confirmation, improving user experience and reducing friction.

Health Assessment

• • Fast cycle time and a single review round indicate a smooth approval process; the small scope of changes minimizes risk to the codebase.

Summary

Remove 'silence this warning' phrasing from instant validation

Aligns error messages with actual severity, improving developer experience and reducing confusion.

Health Assessment

• • Large scope with many test updates, but quick review indicates clear intent and minimal friction.

Summary

Version Packages

This PR updates package versions and changelogs to prepare for a release, enabling automated npm publishing.

Health Assessment

• • Fast cycle with minimal changes, no blockers.

Summary

Automated Package Release

Packages are published to npm with updated changelogs and version numbers, enabling customers to access new features and fixes.

Health Assessment

• • Quick approval with minimal changes; automated release process.

Summary

Backport: Increase Cohere embeddings batch limit

Adds support for larger embedding batches in Amazon Bedrock integration, improving throughput for Cohere models.

Health Assessment

• • Rapid backport with minimal changes and quick merge indicates low complexity and high confidence in the change.

Summary

Increase max embeddings limit for Cohere models

Adjust Amazon Bedrock embedding request limit to match Cohere model capabilities, improving batch processing efficiency.

Health Assessment

• • Fast cycle time and minimal rework indicate a straightforward change with quick review.

Summary

Remove version info from use-ai-sdk skill

Simplifies the use-ai-sdk skill by removing version‑specific content, ensuring evergreen documentation and reducing drift. This improves consistency between the SDK and its documentation.

Health Assessment

• • PR completed quickly with minimal review, indicating a straightforward documentation update.

Summary

Migrate Turbopack build_and_test jobs back to arm64

Increases concurrency for build_and_test jobs, improving CI performance and reducing queue times.

Health Assessment

• • Quick review and merge, minimal changes to CI config, low risk.

Summary

Add Gemini Interactions API support to Vertex provider

Adds support for Gemini's multimodal video generation via Vertex AI, enabling developers to produce video content directly from the Vertex provider. This expands the SDK's multimodal capabilities and simplifies integration for Vertex users.

Health Assessment

• • Rapid approval with minimal iterations indicates high confidence in implementation.

Summary

Backport: fix perplexity chunk schema

Ensures perplexity language model handles optional role and null content, improving robustness.

Health Assessment

• • Fast merge with minimal changes, automated backport, no review needed.

Summary

Automated package version bump and changelog update

This PR updates package.json versions and changelogs for all Vercel AI SDK packages, enabling a coordinated release to npm.

Health Assessment

• • Rapid, single-commit release with no review friction; minimal risk to production.

Summary

Correct example links and typos

This PR cleans up documentation, ensuring accurate links and correct information for developers, improving developer experience and reducing confusion.

Health Assessment

• • The PR had a long review time but minimal changes, indicating low complexity but delayed review.

Summary

Rename cache constants for clarity

Renames internal cache thresholds to more accurately reflect their purpose, reducing confusion and potential bugs in cache handling.

Health Assessment

• • The PR was reviewed and merged within 4.5 hours, with a single commit and no behavior changes, indicating a smooth, low-risk refactor.

Summary

add multi turn examples for gemini omni on google provider

This PR adds multi‑turn text generation and streaming examples for Gemini Omni on Google, enabling developers to quickly prototype and integrate advanced conversational AI into their applications.

Health Assessment

• • Quick turnaround with minimal changes indicates a straightforward addition of example code.

Summary

Upgrade React to latest experimental release

Updates React to the newest experimental version, bringing performance improvements and bug fixes to Next.js applications.

Health Assessment

• • Fast merge with minimal review, indicating high confidence in the upgrade.

Summary

Retry deploy-test logs until markers appear

Ensures deploy tests reliably capture build IDs, preventing intermittent failures and improving test stability.

Health Assessment

• • The PR resolved intermittent test failures by adding a retry loop, but the 4‑day cycle time and 2‑day review delay suggest slow review and potential blocker.

Summary

Fix stale transport in useChat

Ensures that the Chat instance always uses the most recent transport, preventing stale data and improving reliability for users interacting with the chat component. This resolves a bug that could cause incorrect responses when the transport changes during a session.

Health Assessment

• • The PR was reviewed and merged quickly with minimal iterations, indicating a straightforward bug fix with clear tests and examples.

Summary

Replace dynamic import with loadBuiltinModule for diagnostics

Fixes build failures in React Native / Expo apps using Hermes by replacing dynamic import with a Hermes‑safe module loader, ensuring compatibility and preventing compile‑time errors.

Health Assessment

• • Long review cycle indicates possible blocker or low priority; minimal code changes suggest low risk.

Summary

Fix Perplexity chunk schema validation

Fixes streaming errors for Perplexity models, improving reliability for apps using streamText.

Health Assessment

• • Quick fix with minimal changes, resolved validation issue promptly.

Summary

Version Packages

This PR updates package versions for a release, automatically publishing to npm.

Health Assessment

• • Fast merge with minimal changes, typical release PR, no review needed.

Summary

Version Packages

Automated release of package versions, updating changelogs and package.json for multiple SDK packages.

Health Assessment

• • Rapid automated release with minimal human intervention; no review required; low risk.

Summary

Add retry option for failed MCP tool calls

Enables retry logic for MCP tool calls, improving reliability of AI SDK operations.

Health Assessment

• • Fast cycle time and minimal review rounds indicate an efficient backport process. Large code changes were merged quickly, suggesting high confidence in the changes.

Summary

Fix early OTel proxy tracer issue

Ensures cache component prerendering works reliably with OpenTelemetry, preventing runtime errors and improving stability for production deployments.

Health Assessment

Summary

Add retry option for failed MCP tool calls

Adds an optional retry mechanism for MCP tool calls, reducing token waste and improving reliability. This allows agents to automatically recover from transient failures without extra model steps.

Health Assessment

• • The PR required multiple commits to address failing tests and refactor retry logic, indicating moderate complexity. The review cycle was within normal time, but the large code changes suggest careful testing.

Summary

Add gitignore for next-profiles

Prevents large profile files from being committed and causing issues with ecosystem tools.

Health Assessment

• • The PR was merged quickly, indicating a straightforward review process.
• • The changes are focused on a specific issue, reducing the risk of unintended consequences.

Summary

Document xAI responses API default in v7

Adds documentation clarifying that xAI defaults to the responses API in SDK v7, preventing silent behavior changes for users upgrading from v6.

Health Assessment

• • Quick documentation update with minimal review, indicating low complexity and risk.

Summary

Automated package versioning and release

Updates changelogs and package versions for multiple packages, enabling automated npm publishing.

Health Assessment

• • Fast cycle time and minimal changes indicate a smooth release process.

Summary

Version Packages

Automates publishing of updated packages to npm, enabling timely releases.

Health Assessment

• • Release merged quickly with minimal changes, indicating a smooth and efficient release process.

Summary

Version Packages

Automated release of updated package versions and changelogs for the AI SDK packages.

Health Assessment

• • Release PR merged quickly with minimal changes, indicating an efficient release process.

Summary

Health Assessment

Summary

Standardize Insight UI and update docs links

This PR unifies the copy prompt label across the dev overlay and documentation, replaces legacy FixOption components with the new FixCard UI, and updates all Insight error pages with consistent links and related insights navigation, improving developer experience and reducing confusion.

Health Assessment

• • The PR had a quick initial review but required several subsequent commits to address link updates and documentation consistency, indicating moderate iteration but overall efficient resolution.

Summary

Bump eslint with eslint-config-next in codemod

Ensures npm install succeeds by aligning eslint peer dependencies, preventing build failures for projects using Next.js codemod.

Health Assessment

• • Quick review and single commit resolved a peer dependency issue, keeping the change minimal and low risk.

Summary

Add shared harness utilities to reduce duplication

Centralizes sandbox utilities across harness adapters, improving maintainability and reducing drift.

Health Assessment

• • Fast cycle time and minimal net changes indicate low risk and efficient review.

Summary

docs: recommend explicit cacheLife and clarify overriding built-in cache profiles

Adds clear guidance on cacheLife usage in Next.js, improving cache configuration clarity for developers.

Health Assessment

• • Fast cycle time of 23.7 hours and minimal changes indicate a straightforward documentation update with low risk.

Summary

Fix port conflict with vc dev

Fixes intermittent port conflict in Rust dev server, ensuring unique ports and deterministic shutdown, improving reliability for developers.

Health Assessment

• • The PR required multiple iterations and a large code change, indicating significant effort to resolve the port conflict issue. The long cycle time and many commits suggest moderate review friction, but the fix improves developer experience by preventing runtime errors.

Summary

Improve immutable static asset tests

Cleanup and improve tests for static asset handling, ensuring correct path prefix usage.

Health Assessment

• • Fast cycle time and minimal review rounds indicate a straightforward, low-risk test improvement.

Summary

Upgrade Nuxt example to Nuxt 4

Update the Nuxt framework docs example to use Nuxt 4, aligning with official starter template.

Health Assessment

• • The PR had a long cycle time and review period, but only a single commit indicates minimal rework. The large diff suggests significant changes to the example, but the lack of multiple review rounds reduces risk.

Summary

Fix navigation guide highlights and client component page

Updated the instant navigation guide to correct code highlights and add a section on client component pages, improving developer clarity.

Health Assessment

• • Quick documentation update with minimal changes, resolved in under 20 hours.

Summary

Add caveats for bfcacheId and :has

Provides documentation on bfcacheId usage and :has selector caveats, improving developer understanding.

Health Assessment

• • Quick review and minimal changes indicate straightforward documentation update.

Summary

Migrate CI jobs to x86-64 and build native binaries

Enhances the CI pipeline to support both x86-64 and arm architectures, improving build reliability and artifact naming for native binaries.

Health Assessment

• • Quick turnaround with minimal rework, indicating a straightforward CI improvement.

Summary

Add jobs to tests-pass needs list

Adds required jobs to the tests-pass workflow to ensure failures are not skipped. This prevents downstream jobs from incorrectly succeeding when upstream jobs fail.

Health Assessment

• • Quick review and merge with minimal changes; CI workflow updated to enforce failure propagation.

Summary

Bump rust-toolchain sha to fix imposter-commit lint error.

Re-pin dtolnay/rust-toolchain to the current 1.96.1 commit across CI workflows, fixing audit failures and ensuring CI stability.

Health Assessment

• • Quick fix with minimal changes, fast review and merge, low risk.

Summary

Add Gemini Interactions video support

Adds video output support for Gemini Interactions, allowing developers to receive video content directly from the SDK. It also surfaces per-modality token usage and automatically routes Omni model IDs, simplifying integration and providing clearer cost breakdowns.

Health Assessment

• • Fast cycle time and minimal review rounds indicate efficient implementation; large code addition suggests significant feature addition but well-contained.

Summary

Cache exchanged OIDC tokens in memory

Adds an in‑memory cache for OIDC token exchanges, reducing network round‑trips, improving latency, and lowering operational costs for services that repeatedly exchange tokens.

Health Assessment

• • The PR introduced a large in‑memory caching mechanism with extensive documentation and test updates, resulting in a slow 117‑hour cycle time but a quick review due to clear commit and minimal post‑review changes.

Summary

Revert CI test suite runner change

This PR undoes a recent change that moved test suites to x86‑64 runners, restoring the previous runner configuration to avoid CI failures.

Health Assessment

• • Quick revert with minimal changes, resolved within 1.1 hours and no significant review friction.

Summary

Refactor: split validateStaticShell function

Simplifies server rendering logic by extracting static shell validation into a dedicated function, improving code clarity and maintainability. This change reduces potential bugs and eases future enhancements.

Health Assessment

• • The PR took over a week to merge, with a long review period and multiple commits, indicating significant rework and potential complexity.
• • The refactor split validation logic into a separate function, improving maintainability but requiring careful review due to the large diff.

Summary

Move test suites back to x86-64 runners

Mitigates runner pressure by shifting tests to x86-64, reducing build times and resource contention.

Health Assessment

• • Quick change with minimal impact, resolved within minutes.

Summary

Validate Shell Prefetches for Partial Prefetching

Adds robust validation for App Shell prefetches, preventing link data errors and improving developer debugging experience.

Health Assessment

• • Multiple iterations and extensive test changes indicate complex integration; review bot flagged issues; high risk of regressions.

Summary

Add tool filtering to HarnessAgent

Enables fine‑grained control over which tools a harness can use, improving security and compliance for AI runtimes.

Health Assessment

• • Fast cycle time and minimal review comments indicate smooth integration.