Pull Request Explorer

Summary

Handle empty ledger body in mutation-health picker

The nightly mutation health workflow crashed due to empty webhook responses. This change treats empty bodies as zero-row ledger, preventing failures and ensuring new packages are processed correctly.

Health Assessment

• • Rapid fix with minimal changes, resolved critical nightly failure quickly.

Summary

Stabilize AI unit tests against timeouts

Fixes flaky unit tests by warming up heavy loads and increasing test timeouts.

Health Assessment

• • The PR was reviewed and approved quickly, with a small scope and low risk.

Summary

Refactor $evaluateExpression routing via typed-RPC

Enables typed-RPC dispatch for $evaluateExpression, ensuring consistent evaluation across engines and maintaining backward compatibility.

Health Assessment

• • Fast turnaround with a single commit and quick approval indicates a straightforward refactor with minimal rework.

Summary

Refactor core to use typed-RPC dispatcher

Improves the routing of synthetic proxy items through a typed-RPC dispatcher, enhancing the overall performance and maintainability of the system.

Health Assessment

• • The PR was reviewed and merged quickly, indicating a smooth development process. The changes are well-structured and include thorough testing.

Summary

Add human-like code review skill

Enables developers to run a human‑style PR review locally, improving code quality and reviewer efficiency by providing concise, actionable feedback.

Health Assessment

• • The PR was reviewed quickly (0.6h) and merged within 11.5h, with several small commits refining the documentation. AI assistance via Cursor was used to author the skill description.

Summary

Add Postgres connection pool monitoring and recovery

Enables automatic recovery from database connection poisoning, reducing downtime for self‑hosted deployments and improving reliability without requiring orchestrator liveness probes.

Health Assessment

• • Extensive rework with 10 commits after the first review and 5 review rounds indicates a complex change; the large scope and many tests suggest high risk but thorough testing mitigates potential regressions.

Summary

Fix episodic memory scoping for agent channels

Ensures user-specific memory across chat integrations, improving continuity across threads while preserving thread context.

Health Assessment

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

Summary

Fix dropdown menu safety triangle

Restores the dropdown safety triangle that was unintentionally disabled, fixing a UI regression for users.

Health Assessment

• • Rapid, single‑commit fix with AI review, minimal risk and impact.

Summary

Fix external agent channel user ID scoping

Ensures episodic memory is partitioned by platform user rather than thread, improving recall across separate conversation threads while preserving thread-specific history.

Health Assessment

• • The PR was reviewed and merged within 0.9 hours, indicating a smooth process.
• • AI-assisted review contributed to rapid validation and minimal rework.

Summary

Expose mfaEnabled field in Public API

Adds a boolean field to public API user responses, enabling programmatic access to MFA status for security auditing and compliance.

Health Assessment

• • Single commit with no review comments, small scope, low risk of regressions.

Summary

Strip legacy unsupported config before validation

Ensures older agent configurations are sanitized to prevent validation failures, improving reliability of agent configuration persistence.

Health Assessment

• • AI-assisted code and rapid review cycle resolved legacy config validation bug swiftly.

Summary

Move instance AI builds to skill

Routes Instance AI workflow builds through a workflow‑builder skill, gating approvals and preserving artifacts, improving workflow safety and auditability.

Health Assessment

• • PR required extensive rework after the first review, with many commits and a long cycle time, indicating high complexity and potential risk.

Summary

Add sub-agent execution capability

Enables parent agents to delegate bounded subtasks to child agents, improving modularity and allowing linked execution tracking.

Health Assessment

• • Rapid 7‑hour cycle time and AI‑assisted development produced extensive changes and comprehensive tests, indicating a high‑impact feature delivered efficiently.

Summary

Add per-user MCP server connections to instance AI

Users can now link MCP servers to instance AI per user, enabling AI to access tools from those servers with secure authentication, improving workflow automation.

Health Assessment

Summary

Rename OTEL env var to clarify intent

Renames the OpenTelemetry environment variable to a clearer name, ensuring no behavior change but requiring updates to existing configurations.

Health Assessment

• • Quick backport with minimal changes and no review needed, indicating a straightforward rename.

Summary

Rename OTEL env var for clearer intent

Clarifies environment variable naming to ensure production‑only tracing, with no functional change.

Health Assessment

• • Quick approval with minimal review, indicating a straightforward rename with low risk.

Summary

Refactor $fromAI routing via typed-RPC dispatcher

Internal refactor to route $fromAI calls through typed-RPC dispatcher, improving maintainability and ensuring backward compatibility for AI-related expressions.

Health Assessment

• • Rapid merge with minimal review indicates high confidence in changes.
• • Backport to 1.x ensures compatibility across major versions.

Summary

Grant execution owners access to their own redacted data

Enables users who run workflows with private credentials to view their own execution data while keeping it hidden from others, enhancing data visibility and security.

Health Assessment

• • The PR had a long cycle time (171h) and only one review comment, indicating a slow review process and potential bottlenecks. The large code change (377 lines across 16 files) adds complexity and risk.

Summary

Refactor $items() routing via typed-RPC dispatcher

Internal refactor to improve typed-RPC handling of $items() across engines, ensuring consistent behavior and test parity. No user-facing changes, but enhances reliability of expression engine.

Health Assessment

• • Quick merge with minimal review, indicating a straightforward refactor supported by comprehensive tests.

Summary

Refactor $input methods via typed-RPC dispatcher

Enhances input handling by routing $input methods through a typed-RPC dispatcher, improving type safety and reducing runtime errors for users.

Health Assessment

• • The PR was merged within 0.3 hours with no review iterations, indicating a straightforward refactor with minimal risk.

Summary

Restore default build in grind-changed-tests workflow

Fixes CI workflow by reverting to default pnpm build to avoid tsc errors caused by custom build command, ensuring reliable CI runs.

Health Assessment

• • The PR was merged within 18 minutes with a single review comment, indicating a straightforward, low-risk change.

Summary

Add knowledge base to snapshot image

Adds a knowledge base to the Instance AI sandbox image, providing best‑practice templates and unifying skills for improved workflow creation.

Health Assessment

• • Fast review and merge within 3 hours indicates high confidence in code quality; large scope suggests significant feature addition but minimal friction.

Summary

Add datasets field for PR-tier test suite

Enables selective test execution for PR runs, reducing cost while maintaining full nightly coverage.

Health Assessment

• • The PR was reviewed within 34 hours and merged after 60 hours, indicating a moderate review cycle and efficient resolution of issues.

Summary

Fix import workflow subworkflow activation order

Ensures subworkflows are activated before parent workflows during CLI import, preventing failures when using --activeState=fromJson.

Health Assessment

• • Merged within 24 minutes with no review comments, indicating a straightforward, low‑risk bug fix.

Summary

Lock execution row during data-only re-check

Ensures data consistency by locking execution rows during re-checks, preventing race conditions and improving reliability.

Health Assessment

• • Quick review and merge within 3.4 hours indicates smooth process and minimal friction.

Summary

Show eval-mock executions in canvas list

Adds visibility of evaluation mock executions to the n8n canvas, enabling developers to see real execution rows and debug scenarios directly in the UI.

Health Assessment

• • The PR required several rework commits after the initial review, indicating complex changes, but the review was fast and the merge proceeded smoothly.

Summary

Add backend for in‑canvas evaluations wizard

Provides the server logic, DTOs, and validation needed for the new evaluations wizard, enabling users to configure and run evaluation metrics directly from the UI.

Health Assessment

• • The PR had a slow review cycle and multiple review rounds, but only a single commit, indicating a focused change set. The large code addition and backend focus raise the risk level for integration and testing.

Summary

Refactor getNewWorkflowData to workflows API

The getNewWorkflowData helper is moved to the workflows API module, centralizing workflow logic. This refactor improves maintainability and testability while preserving existing functionality.

Health Assessment

• • Fast review and merge with minimal rework, indicating a straightforward refactor with clear intent and thorough testing.

Summary

Fix expression metrics re-registration

Re-register expression metrics after Prometheus registry reset to ensure accurate metrics tracking.

Health Assessment

• • Fast cycle time and immediate review suggest a straightforward and low-risk change.

Summary

Health Assessment

Summary

Fix workflow import order for subworkflow dependencies

Ensures subworkflows are activated before parent workflows during import, preventing failures when using --activeState=fromJson.

Health Assessment

• • The PR was reviewed quickly but required significant rework after refactoring, resulting in a slow overall cycle time yet minimal review friction.

Summary

Declare missing AWS SDK dependencies

Ensures Bedrock nodes have deterministic dependency resolution, preventing runtime failures.

Health Assessment

• • Quick fix with minimal changes, resolved in under an hour with no review friction.

Summary

Replace MCP get_suggested_nodes with best practices tool

Adds richer workflow guidance tool, improving MCP recommendations and user experience.

Health Assessment

• • Long cycle time (~8 days) with multiple review rounds and significant changes to MCP tool, indicating high review friction and moderate risk.

Summary

Migrate ai-utilities tests from Jest to Vitest

Migrates the ai-utilities test suite to Vitest, enabling faster, more efficient testing and aligning with n8n's Jest→Vitest migration strategy.

Health Assessment

• • Fast review and single commit indicate a smooth process; the migration covers a large test suite but encountered no blockers.

Summary

Refactor planner run into briefing and coordinator modules

Reduces technical debt and improves maintainability of the AI orchestration tool, enabling faster feature development and easier debugging.

Health Assessment

• • The PR was reviewed quickly and merged within a day, indicating efficient collaboration.
• • The refactor split a large tool into smaller modules, reducing complexity and improving maintainability.

Summary

Show actions on published version in agent history

Adds action menu to published agent version, enabling revert and unpublish, improving workflow editing.

Health Assessment

• • Quick fix with minimal rework, low risk, fast review cycle.

Summary

Add sub-agent session linkage migration

Adds migration-only schema support for sub-agent execution lineage, enhancing agent thread management.

Health Assessment

• • Quick review and approval process, indicating a well-structured and low-risk PR.

Summary

Improve Instance AI eval reliability and diagnostics

Enhances the stability of AI-generated workflows, reducing failures and providing clearer diagnostics for users.

Health Assessment

• • Rapid AI-assisted review and automated testing enabled quick iteration and a fast 20‑hour cycle time.

Summary

Move CI checkout to useblacksmith/checkout

Updates CI workflows to use a more reliable checkout action, reducing flakiness in Git operations for the release candidate branch.

Health Assessment

• • The PR was merged immediately after creation with no review, indicating a low‑complexity change and high confidence.

Summary

Fix grind workflow to surface vitest errors

Ensures CI accurately reports test failures and provides diagnostics, improving reliability and reducing false positives.

Health Assessment

• • Quick, single‑commit fix with minimal code changes; fast review and merge cycle; improves CI reliability with no new dependencies.

Summary

Release 2.25.1

Minor release with bug fixes and new features for workflow validation and error handling.

Health Assessment

• • Quick release with minimal changes and no significant rework.

Summary

Revert knowledge base snapshot image feature

Rolls back recent feature to maintain stability and prevent regressions.

Health Assessment

• • Quick review and merge within 0.6 hours, indicating minimal friction, but the large code removal suggests a significant impact on the codebase.

Summary

Migrate type-validation throws to UserError

Reclassifies validation failures as user errors, improving error reporting and user experience without changing functionality.

Health Assessment

• • Quick review and single commit indicate low complexity; no new behavior introduced, minimal risk to production.

Summary

Dispatch Expression.evaluate to VM evaluator

Adds VM-based expression evaluation to n8n, enabling isolated execution for improved security and performance, while keeping legacy behavior unchanged.

Health Assessment

• • The PR had a slow review and merge cycle with a large scope, but only a single review round, indicating a complex but ultimately straightforward change.

Summary

Add per-node validation MCP tool

Enables clients to validate individual node configurations quickly, improving workflow reliability and reducing debugging time.

Health Assessment

• • The PR introduced a new MCP tool for node validation, required a single review after a 3‑day wait, and merged after an 8‑day cycle, indicating moderate review friction and a sizable code change.

Summary

Release 2.24.0

This release includes various bug fixes, new features, and improvements to the n8n workflow automation platform.

Health Assessment

• • The PR was merged quickly, indicating a smooth review process.
• • The large scope of the PR may indicate a high level of complexity.

Summary

Fix error workflow trace separation

Ensures error workflows are traced within the same span as the original workflow, improving observability and debugging.

Health Assessment

• • Quick review and minimal changes indicate a low‑risk, low‑friction fix that was merged within five days.

Summary

Migrate @n8n/agents tests from Jest to Vitest

Switches unit test framework to Vitest, improving test performance and consistency across packages.

Health Assessment

• • Fast review and merge with minimal rework, but the migration touches many test files, resulting in a large code change.

Summary

Rebuild isolated-vm for musl libc in Docker

Fixes runtime crash in musl-based Docker images by rebuilding the isolated-vm module, ensuring compatibility and preventing segfaults.

Health Assessment

• • Fast review and merge, but large code changes addressing a critical runtime issue.

Summary

Init VM expression engine in CLI base command

Adds a VM-based expression evaluator to the CLI, improving performance and paving the way for future observability features.

Health Assessment

• • The PR had a 91‑hour cycle time and a 70‑hour review delay, but only one review comment, indicating a large but straightforward change.