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	14283
title	Extract the embedding-dimension helpers out of GoldenPathSynthesizer
state	Closed
labels	airefactoring
assignees	neo-opus-grace
createdAt	Jun 28, 2026, 2:07 AM
updatedAt	Jun 28, 2026, 12:35 PM
githubUrl	https://github.com/neomjs/neo/issues/14283
author	neo-opus-grace
commentsCount	0
parentIssue	14281
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
contentTrust
projected
quarantined	0
signals	[]
blockedBy	[]
blocking	[]
closedAt	Jun 28, 2026, 12:35 PM

Extract the embedding-dimension helpers out of GoldenPathSynthesizer

Closed v13.1.0/archive-v13-1-0-chunk-7 airefactoring

neo-opus-grace commented on Jun 28, 2026, 2:07 AM

Refs #14281

The first (lowest-blast) extraction of the GoldenPathSynthesizer SRP-decomposition — the 3 clearly-misplaced embedding-dimension helpers.

Context

ai/services/graph/GoldenPathSynthesizer.mjs (the 1553-line static grab-bag, #14281) declares 3 top-level export functions that are embedding-dimension utilities, not Golden-Path synthesis: getEmbeddingVectorLength, getEmbeddingModelName, buildEmbeddingDimensionMismatchMessage. V-B-A'd: they have NO external by-name consumers (consumers import the default class); they are used only internally at GPS L1018-1028 (the frontier-embedding dimension-guard); and they have no direct test coverage.

The Fix

Move the 3 functions to a focused module ai/services/graph/embeddingDimension.mjs (pure, no I/O).
GPS imports them back (for the L1018-1028 internal dimension-guard) + re-exports them (API stability — they were public exports; behavior-preserving).
Add focused unit tests (currently untested).

Acceptance Criteria

The 3 helpers live in ai/services/graph/embeddingDimension.mjs; GPS imports + re-exports them (public API stable).
GPS's internal dimension-guard (the L1018-1028 frontier check) still works — behavior-preserving.
Focused unit tests for the 3 helpers: getEmbeddingVectorLength (valid/invalid payloads), getEmbeddingModelName (each provider + default), buildEmbeddingDimensionMismatchMessage (message shape) — green.
node --check clean + the related specs green.

Out of Scope

The other 8 GoldenPathSynthesizer responsibilities (the rest of #14281's leaves).
Any behavior change (this is a pure move + re-export).

Origin Session ID: 090a68e6-1a28-4b20-a5fd-842ebac3e729. Authored by Grace (@neo-opus-grace, Claude Opus 4.8, Claude Code).

tobiu referenced in commit c2eed12 - "refactor(ai): extract the embedding-dimension helpers out of GoldenPathSynthesizer (#14283) (#14285) on Jun 28, 2026, 12:35 PM

tobiu closed this issue on Jun 28, 2026, 12:35 PM

tobiu referenced in commit f66fb7e - "refactor(ai): SRP-decompose GoldenPathSynthesizer — extract the remaining 4 clusters into focused modules (#14281) (#14289) on Jun 28, 2026, 2:31 PM