Pull Request Explorer
| Title | Author | Size | AI | Cycle Time | Review | Merged |
|---|---|---|---|---|---|---|
| fix: preserve provider-executed tool results in length-truncated WorkflowAgent messages | ai-sdk-factory | L | No | 98.2h | 97.9h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| ci: increase OpenAI load time limit to 90ms | gr2m | XS | No | 0.1h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| ci: increase OpenAI load time limit to 90ms | gr2m | XS | No | 0.1h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v5.0] fix: preserve streaming assistant message identity when the chat message list changes | ai-sdk-factory | L | No | 0.2h | 0.2h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: preserve streaming assistant message identity when the chat message list changes | ai-sdk-factory | L | No | 0.3h | 0.3h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v5.0] fix: MCP tool execute incorrectly requires execution options | ai-sdk-factory | S | No | 0.2h | 0.2h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: MCP tool execute incorrectly requires execution options | ai-sdk-factory | S | No | 0.4h | 0.4h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: reject unresolved approved tool calls followed by later context | ai-sdk-factory | L | No | 0.3h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(moonshotai): enable structured outputs for future models | gr2m | M | No | 0.6h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(groq): keep browser search for future model IDs | gr2m | M | No | 0.7h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(cartesia): default future transcription models to streaming | gr2m | M | No | 0.7h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(alibaba): use current video protocols for future Wan models | gr2m | M | No | 0.8h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(mistral): forward reasoning configuration for unknown models | gr2m | M | No | 0.5h | 0.1h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(deepseek): retain reasoning replay for future models | gr2m | M | No | 0.6h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(amazon-bedrock): default future Claude models to supported request fields | gr2m | M | AI | 0.7h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| test(google-vertex): verify forward-compatible Gemini defaults | gr2m | M | No | 0.7h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(workflow): dispatch WorkflowAgent model call telemetry to global integrations from the model step | gr2m | XL | No | 1.6h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: preserve streaming assistant message identity when the chat message list changes | ai-sdk-factory | L | No | 14.7h | 12.8h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: MCP tool execute incorrectly requires execution options | ai-sdk-factory | S | No | 0.6h | 0.6h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: separate retrieved sources from normalized text citations | ai-sdk-factory | XL | No | 240.2h | 0.2h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v5.0] fix: addToolOutput results being dropped for tool calls in earlier messages | ai-sdk-factory | M | No | 1.4h | 1.4h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: addToolOutput results being dropped for tool calls in earlier messages | ai-sdk-factory | M | No | 1.7h | 1.7h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v5.0] fix: preserve web search context across stateless OpenAI Responses steps | ai-sdk-factory | L | AI | 1.6h | 0.4h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: reject unresolved approved tool calls followed by later context | ai-sdk-factory | L | No | 4.6h | 3.4h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| turbopack: report an actionable error when Node.js child processes cannot connect | lukesandberg | L | No | 25.6h | 0.1h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| Turbopack: Ignore realpath errors during module resolution, matching node's behavior | bgw | S | No | 0.7h | 0.7h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| feat(openai-compatible): support uploaded file references via file_id | aayush-kapoor | L | No | 1.0h | 0.0h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| test: add a `vercel` gate condition for Vercel-specific deploy tests | jamiboym | M | No | 43.5h | 41.8h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: prevent OpenCode launch failures with large host tool catalogs | ai-sdk-factory | L | No | 3.2h | 3.2h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v5.0] fix: prevent older overlapping useObject responses from overwriting newer results | ai-sdk-factory | S | No | 0.3h | 0.3h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: preserve the error status of failed native Codex MCP calls | ai-sdk-factory | L | No | 6.1h | 6.1h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| ci: tune test parallelism and avoid recursive cache scans (v5 backport) | gr2m | S | No | 0.2h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| ci: tune test parallelism and avoid recursive cache scans (v6 backport) | gr2m | S | No | 0.2h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| Version Packages | github-actions | L | No | 25.9h | 25.7h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| ci: tune test parallelism and avoid recursive cache scans | gr2m | S | No | 1.9h | - | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: prevent older overlapping useObject responses from overwriting newer results | ai-sdk-factory | S | No | 2.7h | 2.7h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix: addToolOutput results being dropped for tool calls in earlier messages | ai-sdk-factory | M | No | 16.0h | 16.0h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: preserve web search context across stateless OpenAI Responses steps | ai-sdk-factory | L | AI | 2.6h | 2.6h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v5.0] fix: prevent readUIMessageStream from adopting content from an unrelated assistant turn | ai-sdk-factory | M | No | 2.8h | 2.8h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: prevent readUIMessageStream from adopting content from an unrelated assistant turn | ai-sdk-factory | M | No | 2.8h | 2.8h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: round-trip failed Anthropic tool-search results across conversation steps | ai-sdk-factory | M | No | 2.4h | 2.4h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: Anthropic toolChoice 'none' invalidates server-tool history by omitting tool definitions | ai-sdk-factory | S | No | 2.4h | 2.3h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| feat(heygen): backport video provider to AI SDK 6 | ZYPX | XL | No | 0.4h | 0.4h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(ai): prevent global type conflicts across SDK versions | aayush-kapoor | L | No | 24.4h | 0.3h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(ai): prevent global type conflicts across SDK versions | aayush-kapoor | L | No | 24.4h | 0.3h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| fix(ai): prevent global type conflicts across SDK versions | aayush-kapoor | L | No | 24.6h | 0.5h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| Backport: chore(provider/gateway): update gateway model settings files v6 | github-actions | XS | No | 1.0h | 0.8h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| Backport: chore(provider/gateway): update gateway model settings files v5 | github-actions | XS | No | 1.0h | 0.8h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| [v6.0] fix: Handle null error codes in OpenAI Responses streams | ai-sdk-factory | S | No | 4.3h | 4.3h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||
| Reland experimental custom webpack support | sokra | XL | No | 341.3h | 275.7h | Oct 09, 2026 |
|
LLM analysis pending This PR has not been analyzed yet. |
||||||