Pull Request Explorer

Summary

Add episodic memory to agent runtime

Enables agents to retain and recall past interactions, improving contextual decision‑making and user experience.

Health Assessment

• • Fast review and merge (16.5h cycle, 0.1h review time) with minimal rework; large code change but well‑tested and reviewed by AI tool.

Summary

Add episodic memory indexing and reflection

Adds SDK mechanics for episodic entry extraction, ranking, and in-memory store support.

Health Assessment

• • Fast review and merge, with automated review tool used.

Summary

Add AI insights endpoint for eval collections

Introduces a backend endpoint that delivers AI-generated insights for evaluation collections, enabling new UI cards and deterministic fallback logic for reliable results.

Health Assessment

• • The PR experienced a prolonged review cycle with multiple iterations and a substantial code addition, indicating high complexity and potential impact on stability.

Summary

Fix PostHog group identify call after init

Corrects telemetry initialization to prevent errors and ensure accurate user tracking.

Health Assessment

• • Rapid, single-commit fix with AI review confirming no issues; minimal scope and fast cycle time.

Summary

Consolidate memory baseline specs and add no-mcp impact test

Refactors performance tests to isolate memory impact of modules, improving reliability and enabling precise monitoring of memory usage.

Health Assessment

• • Quick review and merge with minimal changes, indicating low complexity and high confidence.

Summary

Rework LangSmith tracing for Instance AI

Improves observability and trace consistency for native agents by normalizing LangSmith data, adding OTel IDs, and enriching metadata for better monitoring.

Health Assessment

• • The PR required extensive refactoring and multiple bug fixes over a long period, indicating significant complexity and potential risk to stability.

Summary

Prevent evaluations tab crash on unsaved workflows

Fixes a crash in the Evaluations tab by ensuring test runs are only fetched after a workflow is saved, improving stability for users creating new workflows.

Health Assessment

• • The PR was reviewed and merged within 19 hours, with minimal code changes and no rework, indicating a straightforward, low-risk fix.

Summary

Fix API key scope for members

Ensures members cannot create API keys with insights:read scope, preventing unintended access.

Health Assessment

• • Quick fix with minimal changes, AI review confirmed no issues.

Summary

Fix scheduled poll isolate acquisition error reporting

Ensures users see errors when scheduled polls fail to acquire isolates, improving visibility and troubleshooting.

Health Assessment

• • Quick fix with minimal changes and single review, indicating low complexity and high confidence.

Summary

Add eval-data row services + eval-setup sub-agent prompt

Adds runtime‑agnostic services for extracting and generating evaluation data rows, plus a system prompt for the eval‑setup sub‑agent, enabling automated evaluation workflows without native agent dependencies.

Health Assessment

• • PR merged after nearly 5 days with a single review round, indicating a straightforward change but a slow review cycle likely due to branch integration delays.

Summary

Add Instance AI evals setup tool

Adds the evals domain tool that turns the eval helpers into an orchestrator-callable setup flow, enabling users to set up evals through the Instance AI tool surface.

Health Assessment

• • The PR has a large scope with 14 files changed, but the review process was relatively smooth with a short time to first review and no major issues reported.

Summary

Set default HITL confirmation timeout to one day

Ensures Instance AI HITL confirmations are bounded by a one‑day timeout, preventing indefinite waits and improving system reliability.

Health Assessment

• • Fast review and merge with no issues, minimal changes, low risk.

Summary

Release 1.123.45

Version bump and dependency updates for stability and security.

Health Assessment

• • Rapid release with no review friction and minimal code changes.

Summary

Reduce Instance AI memory usage

This PR cuts Instance AI’s memory footprint by ~15%, improving performance and lowering operational costs for production workloads.

Health Assessment

• • Fast cycle time (1.9h) with a single review and minimal changes indicates low friction and low risk.

Summary

Release 2.21.5

This release updates version numbers and changelog, delivering the latest bug fixes and stability improvements to users.

Health Assessment

• • Rapid release with minimal changes and a single AI‑assisted review.

Summary

Fix Google Sheets Node column reorder error

Prevents false schema change errors when columns are reordered or inserted, improving reliability of Google Sheets integration.

Health Assessment

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

Summary

Require admin push when migrating legacy environments to projects folder

Adds guard to prevent non‑admin users from pushing during legacy migration, ensuring proper project structure and preventing accidental data loss.

Health Assessment

• • The PR was reviewed quickly with AI assistance, leading to a single review comment and approval. The change is moderate in scope but does not introduce new features, keeping risk low.

Summary

Disable mcp-registry broken E2E test

Temporarily disables a failing end‑to‑end test for the mcp‑registry, preventing CI failures while a proper fix is developed.

Health Assessment

• • Quick approval with minimal review, indicating straightforward change.

Summary

Populate manual user id on webhook execution data path

Ensures manual webhook executions correctly deliver node input/output data to the editor by threading the user identity through the execution path, preventing infinite spinners and data loss.

Health Assessment

• • The PR was reviewed by an AI bot and approved quickly, indicating low complexity.
• • The small code change and clear tests contributed to a smooth merge.

Summary

Bump task-runner-launcher version in runners image

Updates the task-runner-launcher binary to version 1.4.6, eliminating known Go stdlib vulnerabilities in the runners image.

Health Assessment

• • Quick review and merge within 14 hours, minimal changes, low risk security fix.

Summary

Restore spy in stopDuringRun test

Fixes test failures caused by lingering jest spy, ensuring accurate execution updates. Prevents potential runtime errors in production workflows.

Health Assessment

• • Quick fix with minimal changes, resolved in under 15 minutes with a single review comment.

Summary

Harden Slack notifications in release workflows

Improves security by preventing inline interpolation of dynamic values in GitHub Actions, ensuring safer Slack notifications during releases.

Health Assessment

• • Quick review and merge with minimal changes; AI-assisted code likely contributed to speed.

Summary

Stabilize workflow setup credentials state

Fixes freeze in Instance AI workflow setup when selecting Typeform remote form option after credentials configured, ensuring credentials object remains stable and prevents reset loop.

Health Assessment

• • Quick fix with minimal changes, fast review, no blockers.

Summary

Support checking for workflow tags in preExecute hook

Adds ability to query workflow tags during preExecute hook, enabling conditional logic based on tags.

Health Assessment

• • Fast AI-assisted review with no issues, minimal rework, quick merge within 5.5 days.

Summary

Refactor editor to use standalone composable

Improves code organization and reactivity in the workflow editor

Health Assessment

• • The PR has a large scope with 20 files changed, but the review process was relatively smooth with only one commented review

Summary

Fix Instance AI HITL confirmation wait behavior

Ensures HITL confirmations for Instance AI are not prematurely cancelled when users step away, improving reliability of AI workflows.

Health Assessment

• • Quick review and merge within 3.5 hours, minimal rework, indicates straightforward bug fix.

Summary

Fix jsdom XHR leaks in frontend tests

Ensures frontend unit tests run reliably on Node 22 by stubbing XHRs and mocking API calls, eliminating flaky failures.

Health Assessment

• • Quick, targeted fix that resolved flaky test failures on Node 22, with minimal code changes and no impact on production functionality.

Summary

Remove 'Build Agent' button from EmptyStateBuilderPrompt

Eliminate an unused UI element from the editor, simplifying the user experience and reducing confusion.

Health Assessment

• • Fast turnaround with a single review comment and no further changes, indicating a straightforward, low‑risk fix.

Summary

Fix scheduled poll expression isolate acquisition

Ensures polling triggers run correctly under the VM expression engine by acquiring and releasing the expression isolate per poll, preventing runtime errors and maintaining workflow stability.

Health Assessment

• • The PR was merged within 1.4 hours with a single review and no subsequent commits, indicating a straightforward, low-risk bug fix with minimal rework.

Summary

Seeds system credential resolver for private credentials

Adds a system credential resolver that is seeded on startup, hidden from public listings, and protected from modification, enhancing security and workflow configuration.

Health Assessment

• • The PR required multiple iterations after the initial review, indicating complex changes and thorough testing. The AI review helped catch issues quickly, but the team still had to address several failing tests and refactor code before merging.

Summary

Add support for creating and updating workflow groups

Enables users to organize workflow nodes into groups, improving manageability and collaboration.

Health Assessment

• • The PR required extensive rework after a single AI‑driven review, with many commits and a large code change, indicating high complexity and potential risk.

Summary

Add native agent substrate

Adds native checkpoint persistence, memory stores, and native thread/task abstractions to support Instance AI runtime cutover, preparing for full migration to native agents.

Health Assessment

• • Large-scale refactor with extensive migrations and native agent support, requiring multiple rounds of conflict resolution and AI-assisted review, leading to a prolonged cycle time.

Summary

Add eval tools helpers

Adds read‑only utilities that enable automated evaluation of workflows, improving AI integration and reducing manual setup.

Health Assessment

• • Fast review and merge with minimal iterations, indicating clear implementation and thorough testing.

Summary

Resolve backport target branches via remote tags

The backport workflow now correctly resolves target branches using remote tags, preventing fatal errors and silent failures. This ensures backporting of PRs to stable or beta tracks works reliably.

Health Assessment

• • Fast review and merge with minimal changes, indicating a smooth process and low risk.

Summary

Add new AI chat to universal create dropdown

Adds a new AI chat entry to the universal create dropdown, enabling users to start AI conversations directly from the UI and improving workflow efficiency.

Health Assessment

• • Fast cycle time and minimal review comments indicate a straightforward, low-risk feature addition.

Summary

Remove legacy leader election implementation

Eliminates old leader election strategy, consolidating to Lua‑script‑based implementation, improving stability and simplifying code.

Health Assessment

• • Fast cycle time and minimal review indicate a smooth process; AI‑generated summary helped streamline communication; the large refactor carries moderate risk due to extensive code removal.

Summary

Release 2.22.0

Version 2.22.0 release with updated dependencies and documentation.

Health Assessment

• • Fast release with minimal review, indicating stable changes.

Summary

Add reviewer recommendations based on CODEOWNERS

Adds an advisory GitHub Actions workflow that comments on PRs with recommended reviewer teams based on CODEOWNERS, improving review efficiency.

Health Assessment

• • The PR was reviewed and merged within 4 hours, indicating minimal friction. Despite a large addition of 587 lines across 6 files, the change is well-contained to CI tooling, resulting in a low risk profile.

Summary

Hide custom scopes for managed OAuth credentials

This change removes the ability for users to set custom scopes on managed OAuth credentials, preventing scope mismatch issues and ensuring consistent authentication across all customers.

Health Assessment

• • The PR was reviewed quickly with no issues, but required many merges from master, leading to a longer cycle time. The change is a straightforward bugfix with moderate code changes and low risk to production.

Summary

Limit Insights compaction runs

Improves performance and control of insights compaction by adding indexes, guards, and configurable limits, ensuring efficient resource usage.

Health Assessment

• • T
• • h
• • e
• •
• • P
• • R
• •
• • w
• • a
• • s
• •
• • r
• • e
• • s
• • o
• • l
• • v
• • e
• • d
• •
• • q
• • u
• • i
• • c
• • k
• • l
• • y
• •
• • w
• • i
• • t
• • h
• •
• • m
• • i
• • n
• • i
• • m
• • a
• • l
• •
• • r
• • e
• • v
• • i
• • e
• • w
• • ,
• •
• • i
• • n
• • d
• • i
• • c
• • a
• • t
• • i
• • n
• • g
• •
• • a
• •
• • s
• • t
• • r
• • a
• • i
• • g
• • h
• • t
• • f
• • o
• • r
• • w
• • a
• • r
• • d
• •
• • p
• • e
• • r
• • f
• • o
• • r
• • m
• • a
• • n
• • c
• • e
• •
• • f
• • i
• • x
• • .
• •
• • T
• • h
• • e
• •
• • l
• • a
• • r
• • g
• • e
• •
• • l
• • i
• • n
• • e
• •
• • c
• • o
• • u
• • n
• • t
• •
• • r
• • e
• • f
• • l
• • e
• • c
• • t
• • s
• •
• • e
• • x
• • t
• • e
• • n
• • s
• • i
• • v
• • e
• •
• • m
• • i
• • g
• • r
• • a
• • t
• • i
• • o
• • n
• •
• • a
• • n
• • d
• •
• • t
• • e
• • s
• • t
• •
• • u
• • p
• • d
• • a
• • t
• • e
• • s
• • ,
• •
• • b
• • u
• • t
• •
• • t
• • h
• • e
• •
• • s
• • h
• • o
• • r
• • t
• •
• • c
• • y
• • c
• • l
• • e
• •
• • t
• • i
• • m
• • e
• •
• • s
• • u
• • g
• • g
• • e
• • s
• • t
• • s
• •
• • e
• • f
• • f
• • i
• • c
• • i
• • e
• • n
• • t
• •
• • c
• • o
• • l
• • l
• • a
• • b
• • o
• • r
• • a
• • t
• • i
• • o
• • n
• • .

Summary

Wire observational memory runtime

Adds an observation‑memory pipeline to agent execution, improving data consistency and enabling advanced analytics for users.

Health Assessment

• • The PR required extensive review and multiple merge commits, indicating significant rework; the large code churn and many files suggest a high‑impact change that was carefully vetted before merging.

Summary

Migrate @n8n/di from jest to vitest

Switches the testing framework for the dependency injection package to Vitest, improving test performance and compatibility with SWC.

Health Assessment

• • The PR was reviewed quickly but required several commits to resolve issues, indicating moderate iteration but overall low risk.

Summary

Remove user-facing AI research mode

Eliminates the Instance AI research mode UI, simplifying the editor while preserving internal research capabilities for agents.

Health Assessment

• • Fast cycle time and single review comment indicate a smooth refactor with minimal friction.

Summary

Release 2.21.4

This PR releases version 2.21.4, updating package versions and changelog to reflect new fixes and improvements.

Health Assessment

Summary

Release 2.20.12

Patch release with minor fixes and updates.

Health Assessment

• • Quick release with minimal changes and no issues found by the reviewer.

Summary

Restore NDV panel proportions after zoom

Fixes UI distortion after browser zoom, improving user experience and preventing layout issues.

Health Assessment

• • Quick AI-assisted review identified issues, leading to a single round of fixes and a fast merge within 64 hours.

Summary

Fix continueOnFail handling in Send & Wait operations

Corrects API failure routing across eight nodes so that the Continue (using error output) setting works properly, preventing silent success outputs and workflow hangs.

Health Assessment

• • Large scope bugfix with minimal review iterations, indicating efficient resolution but high impact on multiple nodes. AI review contributed to quick validation.

Summary

Revert n8n image to Alpine 3.22

Restores compliance with license policy by reverting to a GPL‑free base image, preventing potential legal issues and ensuring smooth deployment.

Health Assessment

• • Quick turnaround with minimal rework; AI assistance streamlined the revert and smoke test integration.

Summary

Revert n8n image to alpine 3.22

Restores graphicsmagick GPL pin to maintain compatibility with n8n's third-party license policy.

Health Assessment

• • Quick review and merge process, indicating a straightforward fix.

Summary

Optimize CI scripts with caching and shallow checkout

Speeds up CI by caching node modules and reducing checkout depth, improving build times.

Health Assessment

• • Fast review and merge with minimal changes; CI performance gains achieved with low risk.