n8n 2.30.5 - AI Execution Statistics and Alias Fixes

n8n [email protected], published on July 15, 2026, corrects how AI assistant workflow test runs are classified in execution statistics. The patch also retains node aliases in generated AI tool variants and introduces an editor experiment for n8n Credits credential selection. This is a regular release, not a beta or release candidate.

The full release notes and downloads are on the GitHub release page.

n8n now counts AI assistant workflow test runs as manual executions in statistics. Before this patch, that test path was not included with manual runs, so statistics used to inspect interactive workflow development could omit activity initiated through assistant testing. The change is tracked in issue 34153 and implemented by commit e832024.

This matters when an operator compares manual development work with scheduled or event driven production traffic. The corrected classification should make that boundary more accurate without adding a new execution category. It changes the statistical treatment described by the release notes, not the documented execution path itself.

The notes do not mention a backfill, schema migration, or recalculation of historical statistics. Teams that export execution statistics should verify new assistant test runs after the upgrade and should not assume that older records will be rewritten.

For external dashboards, compare like with like on the new version. Run one assistant initiated workflow test and one conventional manual test, then confirm that both enter the manual classification used by the statistics consumer. The notes provide no new metric or category name to map.

The second core fix preserves node aliases when n8n generates AI tool variants. The relevant work appears in issue 34195 and commit 1c73884.

Alias loss is a metadata defect, but it can still make a generated tool variant diverge from the node definition used to create it. This patch keeps that generated representation aligned with the source node alias data. The release notes do not claim changes to node execution, credential resolution, or workflow payloads, so the scope should be treated as alias preservation rather than a broader AI tool rewrite.

A focused validation is enough here. Generate the same AI tool variant before and after the upgrade, inspect the retained aliases, and confirm that an existing workflow still resolves the intended node. That check covers the changed path without turning a small patch into a full workflow migration exercise.

The only feature in this release is an editor experiment for selecting an n8n Credits credential. It is associated with issue 34074 and commit 1cb3d75.

The word experiment sets the useful boundary. The notes do not document a new runtime API, a migration step, a configuration key, or universal availability. Operators should therefore treat it as a narrow editor behavior and avoid changing credential automation based on this release note alone.

For shared n8n instances, this change is mainly relevant to editor regression checks around credential selection. The release provides no evidence of a scheduler, queue, or execution engine change attached to the experiment. The two core fixes remain the parts most likely to affect data collected from workflow development and AI tool generation.