What is the Neural Link?

The Neural Link is a bi-directional bridge that connects AI agents directly to the Neo.mjs runtime. It lets agents inspect the Scene Graph, component state, event listeners, computed styles, and DOM rectangles, and mutate the running application in real time.

Why is Neo.mjs called an Application Engine instead of a framework?

Neo.mjs maintains persistent application objects in a worker-backed Scene Graph instead of compiling application state away into ephemeral DOM nodes. That architecture enables multi-window orchestration, runtime permutation, and deep AI introspection.

What is Context Engineering?

Context Engineering shapes the information and tool environment around AI agents. Neo.mjs implements it through Knowledge Base, Memory Core, GitHub Workflow, and Neural Link MCP servers for frontier harnesses, plus a File System MCP server for internal Neo.ai.Agent local loops.

What is the Neo.mjs Agent OS?

The Neo.mjs Agent OS is the repository Brain: source code and services for Memory Core, Knowledge Base, Active Hybrid GraphRAG, DreamService, Golden Path synthesis, A2A coordination, and Neural Link tooling.

Frontmatter

id	9438
title	TreeStore: Reduce GC pressure and redundant iterations
state	Closed
labels	enhancementaiperformancegrid
assignees	tobiu
createdAt	Mar 11, 2026, 11:09 AM
updatedAt	Mar 11, 2026, 11:14 AM
githubUrl	https://github.com/neomjs/neo/issues/9438
author	tobiu
commentsCount	1
parentIssue	9404
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
contentTrust
projected
quarantined	0
signals	[]
blockedBy	[]
blocking	[]
closedAt	Mar 11, 2026, 11:14 AM

TreeStore: Reduce GC pressure and redundant iterations

Closed v12.1.0/archive-v12-1-0-chunk-1 enhancementaiperformancegrid

tobiu commented on Mar 11, 2026, 11:09 AM

Goal

Optimize TreeStore.mjs to minimize Garbage Collection (GC) pressure and avoid redundant operations, specifically targeting large dataset performance.

Context

During operations like filter, doSort, and bulk expands, the TreeStore can trigger thousands of array allocations and function invocations, significantly stressing the GC.

Acceptance Criteria

Refactor updateSiblingStats() to remove the .filter() array allocation, using a two-pass counter instead.
Hoist activeFilters calculation in filter() to avoid duplicate iteration.
Create a reusable private method #rebuildKeysAndCount() that pre-allocates the _keys array instead of using .map(). Use this method in doSort, filter, clearFilters, expandAll, and collapseAll.
Ensure all tests still pass.

tobiu Mar 11, 2026, 11:13 AM

Input from Gemini 3.1 Pro:

✦ I have applied all the requested optimizations:

Added #rebuildKeysAndCount() to pre-allocate the _keys array, eliminating the .map() overhead and array reallocations in clearFilters, collapseAll, doSort, expandAll, and filter.

Hoisted the activeFilters array evaluation to the top of filter(), avoiding a duplicate iteration.

Completely rewrote updateSiblingStats() to use a two-pass counting and indexing loop, eliminating the .filter() call that was previously creating thousands of short-lived arrays.

All unit tests pass successfully. The commit has been pushed. Closing this optimization sub-ticket.