Pull Request Explorer

Summary

Fix Salesforce JWT upload and error handling

Corrects document upload failures and surfaces Salesforce error messages for JWT authentication, improving reliability for users.

Health Assessment

• • Rapid turnaround with minimal rework indicates a straightforward bug fix.

Summary

Replace deprecated Azure dedicated host SKU

Nightly benchmark runs now succeed by switching to a supported Azure host SKU, preventing provisioning failures and maintaining performance baselines.

Health Assessment

• • Quick AI-assisted fix with minimal changes, no rework, fast review, low risk to production

Summary

Centralize zod CJS-pin alias in Vitest config

Reduces duplication and fragility in test configurations, ensuring consistent test behavior across packages.

Health Assessment

• • Quick review and merge with minimal changes, indicating a straightforward refactor with low risk.

Summary

Stub node-icon SVG imports in Vitest config

Fixes flaky test failures by stubbing SVG imports, ensuring stable test runs without affecting production rendering.

Health Assessment

• • Quick, single-commit fix that resolved intermittent test failures without impacting production behavior.

Summary

Fix Salesforce Node JWT upload and errors

Corrects document upload failures and surfaces Salesforce error messages for JWT authentication, improving reliability for users.

Health Assessment

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

Summary

Add durable scheduler retention pruning

Enables automatic pruning of finished scheduler tasks, preventing unbounded growth of the task table and reducing storage costs.

Health Assessment

• • The PR was reviewed and merged within 2.1 hours with minimal rework, indicating high confidence in the implementation. The large code change was accepted quickly, suggesting the changes were well-structured and well-tested.

Summary

Fix Salesforce Node JWT upload and error surface

Corrects document upload failures and exposes Salesforce error messages for JWT authentication, improving reliability and debugging for users.

Health Assessment

• • Quick review and single commit fix after initial review indicates efficient process; moderate scope but resolved within a day.

Summary

Pin langchain core packages to prevent mismatches

Ensures consistent dependency versions for langchain, preventing startup crashes and simplifying debugging. This reduces runtime errors and improves stability for users.

Health Assessment

• • Fast cycle with minimal review rounds; quick resolution of dependency issue.

Summary

Add certificate authentication to Microsoft Entra Service Principal credential

Adds certificate support to Microsoft Entra Service Principal credentials, enabling secure client assertion authentication and expanding authentication options for existing nodes without code changes.

Health Assessment

• • The PR introduced certificate authentication for Microsoft Entra Service Principal credentials, expanding authentication options. The review was quick but required multiple integration merges, indicating moderate integration friction but no major blockers.

Summary

Load MCP tools from server for instance AI

Ensures instance AI uses up‑to‑date MCP tools, improving accuracy and giving users control over tool selection.

Health Assessment

• • PR required extensive rework after the first review, with multiple commits and review rounds, indicating a complex change that impacted both backend and frontend components. The fast initial review time was offset by high overall friction and a large scope, raising the risk level for integration and testing.

Summary

Overhaul knowledge base retrieval

Improves agent sandbox reliability, reduces search timeouts, enforces storage limits, and cleans up orphaned sandboxes, enhancing overall agent performance and stability.

Health Assessment

• • PR involved extensive rework with 10 commits and large line changes, indicating complex modifications; review time was moderate but required multiple iterations, raising risk for regressions.

Summary

Fix auto‑open template credential modal

Ensures the credential setup modal reliably opens for template imports, improving user experience and eliminating flaky tests.

Health Assessment

• • Quick review and minimal changes led to a smooth merge, but the overall cycle time was extended due to scheduling or other unrelated delays.

Summary

Streamline custom role scope handling

Ensures consistent, validated custom role scopes across UI and backend, reducing errors and improving security.

Health Assessment

• • Fast review and single round of comments suggest confidence, but the large code churn indicates substantial refactoring effort.

Summary

Fix missing intermediate steps in AI Agent node

The v3 AI Agent node now always returns an intermediateSteps array when enabled, matching v2 behavior. This prevents downstream nodes from failing when the model answers directly without tool usage.

Health Assessment

• • Single commit with quick review; low complexity; PR merged after ~4 days.

Summary

Replace daytona volume with BinaryDataService

Replaces the legacy Daytona volume storage with the new BinaryDataService, streamlining agent data handling and improving performance and reliability.

Health Assessment

• • The PR was reviewed and merged within minutes, indicating a smooth review process and minimal friction.
• • Despite a large code change, the lack of back‑and‑forth discussion suggests confidence in the implementation.

Summary

Add multi‑main support for Instance AI

Enables Instance AI to operate reliably across multiple main processes, improving scalability and resilience for enterprise deployments.

Health Assessment

• • The PR required extensive rework and multiple review rounds, indicating high complexity. Large code churn and cross‑process logic changes elevate risk for production stability.

Summary

Fix JSON expression marker stripping

Ensures JSON body fields correctly parse after toggling expression mode, preventing parse errors.

Health Assessment

• • Long review cycle but minimal changes; likely due to backlog or bot workflow.

Summary

Return empty form values when capturing browser secrets

Fixes a bug where empty form values were incorrectly captured as text, ensuring accurate secret capture.

Health Assessment

• • Quick resolution with minimal rework; AI-generated summary streamlined communication.

Summary

Release 2.28.7

Patch release that pins langgraph dependencies to prevent broken npm installs, improving stability for users.

Health Assessment

• • Fast release with minimal changes and no review friction.

Summary

Release 2.29.6

Bug fixes to core dependencies to prevent broken npm installs.

Health Assessment

• • Fast release with minimal changes and no review needed.

Summary

Move createDeferredPromise into @n8n/utils

Pure relocation of utility function to central utils package, no behavior change.

Health Assessment

• • The PR involved a single commit and a single review comment, indicating a straightforward refactor with minimal friction.

Summary

Pin langgraph packages to prevent npm install failures

Ensures stable npm installs by pinning langgraph dependencies, preventing runtime errors for users.

Health Assessment

• • Quick fix with minimal changes; AI-assisted commit; backport to release candidate; no major rework.

Summary

Pin langgraph packages to prevent npm install errors

Ensures stable npm installs for n8n 2.28.x by pinning langgraph dependencies, preventing runtime errors for users.

Health Assessment

• • Single commit, no review delays, quick merge indicates low complexity.

Summary

Pin langgraph packages to fix npm install errors

This PR pins specific langgraph dependencies to stable versions, preventing runtime errors for npm consumers and ensuring consistent package resolution across environments.

Health Assessment

• • Fast review and minimal changes indicate low risk; the fix addresses a critical dependency issue with minimal code churn.

Summary

Add plan-rejection evaluation and harden multi-turn consistency

Adds a regression guard for HITL plan-rejection handling, ensuring agents honor rejections and revise plans accordingly, improving reliability of AI-driven workflow planning.

Health Assessment

• • Single review with minimal comments; long cycle time likely due to scheduling, but overall low risk.

Summary

Update node popularity data

Automated update of node popularity data for sorting nodes in the node creator panel, reflecting current usage patterns.

Health Assessment

• • Automated weekly update with zero review time, minimal risk.

Summary

Extract reusable credential form controller

Centralizes credential form logic, enabling reuse across UI surfaces and reducing code duplication.

Health Assessment

• • Quick review and single commit, minimal rework, indicates a straightforward refactor with clear test coverage.

Summary

Add durable scheduler materialization sweep

Precomputes scheduled tasks to improve scheduling performance and reliability, reducing runtime overhead and ensuring consistent timing across instances.

Health Assessment

• • Fast 20‑hour cycle with a single AI review and no rework, but the change spans 20 files and nearly 2,000 lines, indicating a substantial backend refactor that is well tested.

Summary

Add breaking change rule for multi-output data

Introduces a rule to detect breaking changes when nodes produce multiple outputs, improving workflow stability.

Health Assessment

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

Summary

Exclude streaming endpoints from Sentry tracing

Prevents long‑lived streaming endpoints from skewing Sentry latency metrics, improving monitoring accuracy.

Health Assessment

• • Fast review and merge with minimal changes, indicating low complexity and risk.

Summary

Deprecate $getPairedItem in the expression editor

Deprecates the undocumented $getPairedItem helper in the frontend expression editor by surfacing an error, guiding users away from it while preserving backend execution behavior.

Health Assessment

• • Fast cycle time (2.1h) and minimal review rounds indicate a smooth process; AI review by cubic likely contributed to the quick approval.

Summary

Improve Instance AI workflow reliability

Enhances the stability of AI-generated workflows, reducing intermittent failures and improving user experience.

Health Assessment

• • Rapid review and AI-assisted feedback enabled a concise, high‑impact fix that tightened workflow reliability across the platform.

Summary

Add durable scheduler executor

Adds a durable scheduler executor that claims, dispatches, and leases scheduled tasks, improving reliability and concurrency for scheduled workflows.

Health Assessment

• • Fast review and merge with minimal rework; large code addition is well-tested, indicating low risk to production.

Summary

Refactor Instance AI: Remove Dead Code

Clean up unused code and dependencies in the Instance AI package, improving maintainability and reducing bundle size.

Health Assessment

• • The PR was merged within 1.3 hours with minimal review comments, indicating a straightforward cleanup.
• • The large line change count reflects extensive removal of dead code but no functional impact.

Summary

Gate tag write actions behind RBAC scopes

Prevents unauthorized users from seeing or attempting tag modifications, improving security and UI clarity.

Health Assessment

• • The PR fixed a security UI bug, required multiple iterations after a quick review, and touched many files, indicating moderate review friction but a normal overall cycle time.

Summary

Fix MCP Client Node token handling

Ensures MCP client does not send undefined tokens and includes resource URL in OAuth requests, preventing 500 errors and improving authentication reliability.

Health Assessment

• • The PR required a review request and multiple iterations, indicating moderate complexity and careful validation of authentication logic.

Summary

Fix AI snapshot empty tree rendering

Ensures assistant messages correctly display tool calls, improving user experience by preventing empty bubbles when snapshots are missing.

Health Assessment

• • Fast turnaround with a single commit and minimal review comments indicates a straightforward, low-risk bug fix.

Summary

Make agent execution project-scoped

This PR removes the userId dependency for agent execution, enabling scheduled agent tasks and integration reconnects in team projects, thereby improving reliability and security for workflow automation.

Health Assessment

• • Fast cycle time and quick review indicate an efficient process; the large number of lines changed reflects a significant impact on agent execution logic.

Summary

Improve OTEL error type handling

Enhances error diagnostics for OpenTelemetry traces, ensuring accurate error naming and fallback for unknown errors, improving observability and debugging.

Health Assessment

• • Quick turnaround with minimal rework; AI-generated summary and review helped streamline the process.

Summary

Broaden secret detection in browser snapshots

Enhances browser snapshot redaction to detect more secret patterns, reducing risk of leaking sensitive data.

Health Assessment

• • Fast cycle time of 2.8 hours and a single review round indicate a smooth process with minimal friction. The change is moderate in scope but carries low risk.

Summary

Add public API endpoints for evaluation test runs

Enables external reporting tools to poll evaluation test run status, metrics, and case results via authenticated, cursor‑paginated endpoints.

Health Assessment

• • Fast cycle time and minimal review iterations suggest a well‑scoped, efficiently implemented feature with low risk of regressions.

Summary

Export contained workflows when exporting a folder

Adds ability to export workflows inside folders when exporting a package, improving export completeness and preventing missing workflows.

Health Assessment

• • PR was merged in under 21 hours with a single review and no rework, indicating efficient implementation and thorough testing.

Summary

Persist AI Assistant domain-access approvals

Ensures users no longer need to re‑approve domains after restarts or across multiple instances, improving experience and reducing friction.

Health Assessment

• • Fast review cycle with a single iteration; the change fixes a user‑experience issue by persisting approvals across restarts.

Summary

Shorten AI thread auto-removal to 30 days

Reduces the default retention period for AI assistant threads to 30 days, preventing database growth and improving storage efficiency.

Health Assessment

• • Fast turnaround with minimal changes; backport completed quickly with no review needed.

Summary

Configure Instance AI providers via environment variables only

Enables self‑hosted deployments to configure AI providers solely through environment variables, removing UI credential handling and simplifying setup while maintaining security and consistency across deployments.

Health Assessment

• • Merged in under 20 minutes with a single commit and no review comments, indicating a straightforward change with minimal friction.

Summary

Update README with new product positioning

Updated product positioning text in README to better reflect product info, improving clarity for users.

Health Assessment

• • Quick review and merge within 1 hour, minimal changes, no blockers.

Summary

Fix instance AI title refinement logic

Ensures AI-generated thread titles are accurate by stripping injected metadata from user messages, preventing misleading titles and improving user experience.

Health Assessment

• • The PR was merged within 3.8 hours with a single review and no rework, indicating a smooth, low‑risk fix.

Summary

Shorten default AI assistant thread auto-removal to 30 days

Reduce default thread TTL to 30 days to prevent database growth, still configurable via environment variable.

Health Assessment

• • Quick review with no issues, minimal changes, low risk

Summary

Configure Instance AI providers via environment variables

By hiding model, search, sandbox, and advanced sections, the system now relies solely on environment variables for AI provider configuration, enhancing security and simplifying deployment. This reduces the risk of exposing credentials through the UI and streamlines configuration for self-hosted instances.

Health Assessment

• • The PR was reviewed and merged within 19 hours with no comments, indicating high confidence and low risk. The change involves a moderate codebase shift but no complex dependencies or extensive rework.

Summary

Unify Microsoft Service Principal target validation

Ensures consistent and secure validation of Service Principal credentials across Microsoft nodes, improving reliability and preventing mis‑targeting errors for end users.

Health Assessment

• • Fast review and approval with minimal rework, but the change spans many files and lines, indicating a substantial backend refactor that could impact multiple integrations.