Pull Request Explorer

Summary

Fix inspector shimmer to label span

Corrects UI shimmer effect to target only label text, improving visual clarity and performance.

Health Assessment

• • Rapid iteration with AI‑assisted review enabled quick resolution of a UI bug across many components.

Summary

Add document_likes table for workspace likes

Introduces a table to store likes on workspace documents, enabling like counts and user-specific like tracking for the application.

Health Assessment

• • The PR was merged quickly (17h) but required several commits after the initial review, indicating complex logic around document transfer and likes. Extensive test coverage helped mitigate risk despite the large code changes.

Summary

Refactor aiAgent: extract InterventionController and threadRunHooks

Improves maintainability and testability of the aiAgent service by separating concerns into dedicated controller and hook modules.

Health Assessment

• • Single commit refactor with quick merge, minimal review friction.

Summary

Scroll, group and sweep acceptance list

Adds infinite scrolling, grouping, multi‑select batch actions, and an improved rename modal to the acceptance list, boosting user productivity and UI clarity.

Health Assessment

• • The PR introduced a large UI feature with significant code changes, required multiple AI‑assisted review iterations, and was merged within a day, indicating moderate review friction but high implementation risk due to scope.

Summary

Correct observability test fixture cast

Fixes a type casting issue in observability tests, ensuring code quality checks pass and preventing potential runtime errors.

Health Assessment

• • Quick fix with minimal changes, no review needed, merged within minutes.

Summary

Coalesce Gateway tool state updates for performance

This change reduces redundant full-message reconciliations, speeding up tool output streams. It does so without affecting the user interface.

Health Assessment

• • Fast cycle time and minimal changes indicate a smooth review process with low risk.

Summary

Fix repeated tool call loops in agent runtime

Prevents infinite loops of tool calls, improving stability and user experience.

Health Assessment

• • Quick resolution with minimal rework, indicating clear issue and efficient review.

Summary

Hide dates on historical brief items

This change removes redundant dates from older brief items, improving readability and reducing visual clutter for users.

Health Assessment

• • Quick fix with minimal changes, reviewed by AI, merged within hours.

Summary

Center reject modal instead of pinning left

Fixes UI alignment for reject modal in acceptance checks, ensuring it is centered and consistent with portal header, improving user experience.

Health Assessment

• • Quick fix with minimal changes, merged within 15 hours, indicating smooth review and low complexity.

Summary

Fix omitted todo operation type inference

Ensures updateTodos operations are validated and errors are surfaced, preventing silent failures and repeated calls.

Health Assessment

• • Quick fix with minimal rework, AI-assisted code generation improved reliability.

Summary

Keep running turn's process collapsed

Improves chat UI by collapsing tool outputs during agent turns, reducing clutter and enhancing user experience.

Health Assessment

• • The PR received a quick review with no major rework and merged within 20 hours, indicating a low‑risk, straightforward UI enhancement.

Summary

Add parent folder support for agent documents

Enables hierarchical organization of agent documents by allowing a parent folder to be specified during creation, improving structure and preventing filename collisions.

Health Assessment

• • Quick single-commit PR with minimal review, indicating a straightforward change.

Summary

Add graph read limits, lifecycle tracing, deadline budget

Improves performance and reliability of goal tracking by limiting event payload, adding lifecycle events, and enforcing deadlines.

Health Assessment

• • Fast review cycle (0.1h) and minimal rework indicate a smooth process; automated AI review helped catch issues early.

Summary

Fix Kimi Code render registry type

Aligns Kimi Code render map with builtin registry, ensuring type safety and compatibility.

Health Assessment

• • Quick fix with minimal changes, merged within minutes, indicating low complexity and high confidence.

Summary

Unpin @ant-design/icons after 6.3.4 fix

Update dependency to allow newer icon package, preventing install failures. Ensures fresh installs resolve the fixed icon package, eliminating build errors.

Health Assessment

• • Q
• • u
• • i
• • c
• • k
• •
• • f
• • i
• • x
• •
• • w
• • i
• • t
• • h
• •
• • m
• • i
• • n
• • i
• • m
• • a
• • l
• •
• • c
• • h
• • a
• • n
• • g
• • e
• • s
• • ,
• •
• • m
• • e
• • r
• • g
• • e
• • d
• •
• • w
• • i
• • t
• • h
• • i
• • n
• •
• • 1
• • 6
• •
• • h
• • o
• • u
• • r
• • s
• • ,
• •
• • i
• • n
• • d
• • i
• • c
• • a
• • t
• • i
• • n
• • g
• •
• • s
• • m
• • o
• • o
• • t
• • h
• •
• • r
• • e
• • v
• • i
• • e
• • w
• •
• • a
• • n
• • d
• •
• • l
• • o
• • w
• •
• • r
• • i
• • s
• • k
• • .

Summary

Harden sync acknowledgements and trace export

Improves reliability of full-text search sync by retrying deadlocks and enhances observability by allowing multiple trace backends.

Health Assessment

• • Quick fix with minimal changes, no major rework.

Summary

Add Kimi Code tool inspectors and renders

Provides dedicated inspector and render surfaces for Kimi Code tools, improving debugging and visualization for developers.

Health Assessment

• • Fast review and merge with minimal rework; single commit and quick approval indicate a straightforward, low-risk enhancement.

Summary

Parallel task scheduling, trace fixes, CLI update

Improves goal execution efficiency, fixes trace errors, and updates the CLI for a smoother user experience.

Health Assessment

• • The PR required multiple commits after the first review and two automated review rounds, indicating significant rework, but was merged within 10 hours. The large code churn and many changes suggest a high-risk, high-effort effort that was successfully completed.

Summary

Plan exploration structure before executing goals

Adds a planning step to goal creation, enabling decomposition into multiple tasks and improving graph visualization for better user experience.

Health Assessment

• • The PR underwent extensive iteration after a single review, indicating complex changes, but was merged quickly due to efficient resolution.

Summary

Add keyset-paged list read and batch status/delete writes

Enables efficient paginated reads and batch updates for acceptance records, improving UI responsiveness and data integrity.

Health Assessment

• • Fast review and single commit indicate a smooth process; the large code change is well-tested and unlikely to introduce regressions.

Summary

Refactor AiAgent: Extract helpers and types

This refactor splits the large AiAgentService into smaller modules, improving code readability and maintainability. It preserves the public API, ensuring no impact on existing functionality.

Health Assessment

• • Quick single-commit refactor with no test changes and minimal code movement, indicating low risk and efficient resolution.

Summary

Hide duplicate check on stale model row

Improves UI clarity by removing conflicting checkmark and aligning switch for disabled model rows, enhancing user experience.

Health Assessment

• • Quick merge with minimal changes indicates low complexity and high confidence in UI improvement.

Summary

Integrate Mesurer inspector into DevDock

Adds a Mesurer visual inspector toggle to DevDock, enabling developers to inspect UI components directly within the app and improving debugging efficiency.

Health Assessment

• • The PR was reviewed and merged within 18 minutes, indicating a smooth, low‑friction process with minimal rework.

Summary

Fix bound long Elasticsearch queries

Improves search reliability by limiting query size, preserving data integrity, and adding robust error handling.

Health Assessment

• • Rapid review and merge (0.1h cycle time) indicates high confidence and minimal rework; the large code change was accepted quickly with no additional iterations.

Summary

Add one‑click reconnect for offline devices

Enables users to quickly reconnect offline devices with a single action, improving device management and reducing downtime.

Health Assessment

• • Fast cycle time and minimal rework indicate a well‑scoped change; AI review helped streamline feedback.

Summary

Merge font size into Font section

Improves user experience by consolidating font settings, simplifying the UI, and enabling live preview of font changes.

Health Assessment

• • Single commit refactor with quick merge, minimal review, low risk.

Summary

Fix Mermaid preview overflow in chat settings

Ensures Mermaid diagrams display correctly by removing fixed height and adding responsive width in chat appearance settings.

Health Assessment

• • Quick UI fix with minimal changes, no tests required, resolved overflow issue in chat settings.

Summary

Add support for additional trusted origins

Introduces an environment variable that lets users merge extra trusted origins into the authentication configuration, improving flexibility for self‑hosted deployments.

Health Assessment

• • The PR closed in under three hours with a single commit and minimal review, indicating a straightforward change with low complexity and risk.

Summary

Add bot channel filter to topics page

Users can now filter agent topics by the bot channel they were sent from, improving search and organization.

Health Assessment

• • The PR had a slow cycle time of nearly three days but received a quick initial review, with moderate rework across three commits to refine the filter UI.

Summary

Add branch name copy action

Enables users to copy branch names directly from the UI, improving workflow efficiency.

Health Assessment

• • Fast cycle time and single AI‑reviewed comment indicate a smooth, low‑risk change.

Summary

Add goal detail drill‑down via portal

Enables users to inspect goal details and metrics directly in the portal, improving transparency and reducing manual steps.

Health Assessment

• • Rapid approval with minimal rework indicates high confidence in implementation; large scope suggests significant feature addition.

Summary

Pin @ant-design/icons to 6.3.2

Ensures stable icon imports by pinning the dependency to a known working version, preventing build failures in the SPA.

Health Assessment

• • Quick fix with minimal changes, merged within 2 hours, indicating low complexity and high confidence.

Summary

Fix OIDC to accept apex desktop redirects

Enables desktop OIDC clients to use apex gateway URLs, maintaining compatibility while allowing new entrypoint without changing issuer or app URL.

Health Assessment

• • Quick fix with minimal changes, merged within 1.3 hours, indicating efficient review and low complexity.

Summary

Pin @ant-design/icons to 6.3.2

Fixes build failures caused by incompatible icon dependency, ensuring stable deployments.

Health Assessment

• • Resolved quickly with a single review and minimal code change.

Summary

Remove dead modal header overrides

Eliminates unnecessary header styling for skill store modals, improving UI consistency and reducing code clutter.

Health Assessment

• • Fast cycle time and minimal review rounds indicate a straightforward, low-risk refactor. AI-assisted review provided concise feedback, speeding up the process.

Summary

Show agent avatar on scheduled task notifications

Adds the executing agent’s avatar to scheduled task notifications, giving users immediate visual context of who performed the task and reducing confusion in the inbox.

Health Assessment

• • The PR was merged within 1.4 hours with a single commit and no review comments, indicating a straightforward, low-risk change.

Summary

Restore node glyph for goal show after rename

Fixes a CLI display bug where task rows showed undefined glyph after a node kind rename, improving user experience and preventing confusion.

Health Assessment

• • Quick, single-commit fix with fast review turnaround and minimal code changes, indicating low complexity and risk.

Summary

Restore draft text label on topic rows

Reverts a recent change to bring back the WeChat‑style draft prefix, improving user clarity on topic rows.

Health Assessment

• • Quick revert with minimal changes, no new tests, low risk.

Summary

Fix sidebar highlight, unify skeleton, move scope switch

Improves resource page UX by ensuring correct file selection, consistent loading skeletons, and a cleaner header scope toggle, enhancing navigation clarity.

Health Assessment

• • Quick turnaround with minimal rework; single review comment resolved in one commit.

Summary

Refactor acceptance report output to temp root

Acceptance report rounds are now scaffolded into a temporary directory instead of the repository, improving isolation and reducing clutter.

Health Assessment

• • The PR was reviewed and merged within 17 hours, with a single review comment and no additional commits, indicating a smooth, low‑risk change.

Summary

Add eval case capture from conversation

Enables developers to capture a conversation turn as an eval test case, improving regression testing and data quality.

Health Assessment

• • The PR was merged in under 16 hours with a single review, indicating rapid iteration and high confidence in the changes.

Summary

Fix file search and staged deletions

Improves file search handling, ensuring accurate results for filtered queries and staged deletions, enhancing user experience in file management.

Health Assessment

• • Rapid review and minimal rework indicate a straightforward bug fix with clear test coverage.

Summary

Refactor Home Recents Store for Local-First

Improves startup speed and offline usability by caching recent items locally, enhancing user experience and reducing server load.

Health Assessment

• • Fast cycle time and AI-assisted review enabled rapid iteration, but the large scope and multiple post-review commits suggest moderate risk of regressions.

Summary

Add Factory Droid ACP integration

Enables Factory Droid as a heterogeneous local CLI agent, expanding device integration and user control.

Health Assessment

• • The PR received a quick automated review and only two follow‑up commits, indicating low review friction, but the large number of added lines and files suggests a high‑impact change that carries moderate risk.

Summary

Accelerate and Harden Elasticsearch Reindex Recovery

Improves migration reliability and performance for high‑volume data, reducing downtime and memory usage during production reindexing. Ensures safe, parallel recovery and compatibility with Serverless deployments.

Health Assessment

• • Rapid 7‑hour cycle time with multiple commits shows active iteration, but the large code changes (≈840 lines across 20 files) indicate a high‑impact update that required careful coordination.

Summary

Add goal trajectory recording and replay

Enables recording of full goal execution paths and provides CLI tools to view and replay them, improving observability and debugging for agents.

Health Assessment

• • The PR was reviewed quickly (0.1h) and merged after a single review round, but it introduces a large codebase change (~40k lines) that adds new data structures and CLI commands, representing a high-risk feature addition.

Summary

Add structured, git-invisible home for report rounds

Organizes acceptance test outputs into a hidden directory, preventing clutter in git history and improving traceability for reviewers.

Health Assessment

• • Fast review cycle and minimal rework indicate a smooth integration; documentation and tests support the change.

Summary

Fix task assignment and list N+1 issue

Improves task list performance by eliminating N+1 queries and separating member and agent assignments, reducing API load and enhancing user experience.

Health Assessment

• • The PR underwent rapid, iterative changes with frequent automated review comments, indicating a complex refactor that required multiple adjustments before final approval.

Summary

Add test case page, addressable without run

Provides a dedicated page for individual test cases, enabling direct access and editing, improving usability and traceability.

Health Assessment

• • The PR introduced a large number of changes across both frontend and backend, but received rapid review and a single round of feedback, indicating a well‑structured change with minimal friction.

Summary

Show credits for heterogeneous API agents

Adds a credits display for heterogeneous API agents, improving transparency and user awareness of usage limits.

Health Assessment

• • Fast cycle time and minimal rework indicate smooth integration and low risk to the codebase.