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

title	chore(ai): clean residual memory-core comments (#11922)
author	neo-gpt
state	Merged
createdAt	May 28, 2026, 8:01 AM
updatedAt	May 28, 2026, 9:35 AM
closedAt	May 28, 2026, 9:35 AM
mergedAt	May 28, 2026, 9:35 AM
branches	`dev` ← `codex/11922-memory-core-residual-comments`
url	https://github.com/neomjs/neo/pull/12127

Merged

neo-gpt commented on May 28, 2026, 8:01 AM

Authored by GPT-5 (Codex Desktop). Session f2619b83-0862-4a38-b5fa-11bcb9ee646d.

FAIR-band: over-target [14/30] - taking this lane despite over-target because it is a tiny, already-assigned #11922 residual cleanup/closeout batch with no runtime surface and clears owned v13 cleanup debt.

Refs #11922 Related: #11912

Removes residual source-comment archaeology from two Memory Core files by rewriting historical phase/ticket markers into stable runtime-contract wording. This keeps the cleanup aligned with the #11912 source-comment contract while preserving behavior.

Evidence: L1 (static source-comment cleanup plus syntax/diff checks) -> L1 required (comment/JSDoc-only cleanup ACs). No residuals for this batch.

Deltas from ticket

Narrowed to PermissionService.mjs and CoalescingEngineService.mjs residual anchors discovered after earlier #11922 merges.
Preserved runtime behavior and runtime strings; no generated/config surfaces touched.

Test Evidence

git grep -n -E "#[0-9]{4,}|PR #[0-9]{4,}|AC[0-9]+|Cycle [0-9]|cycle-[0-9]|Discussion #[0-9]{4,}|phase [0-9]" origin/dev -- ai/services/memory-core/PermissionService.mjs ai/services/memory-core/CoalescingEngineService.mjs showed 7 pre-existing residual anchors in touched files.
rg -n "#[0-9]{4,}|PR #[0-9]{4,}|AC[0-9]+|Cycle [0-9]|cycle-[0-9]|Discussion #[0-9]{4,}|phase [0-9]" ai/services/memory-core/PermissionService.mjs ai/services/memory-core/CoalescingEngineService.mjs returned no matches after cleanup.
node --check ai/services/memory-core/PermissionService.mjs passed.
node --check ai/services/memory-core/CoalescingEngineService.mjs passed.
git diff --check origin/dev...HEAD passed.
Branch freshness: merge-base HEAD origin/dev == origin/dev; branch history contains only f5ccfbd2f chore(ai): clean residual memory-core comments (#11922).

Post-Merge Validation

Re-run the #11922 focused diagnostic when the broader Memory Core group is ready for issue closeout.

Commit

f5ccfbd2f - chore(ai): clean residual memory-core comments (#11922)

neo-opus-ada APPROVED reviewed on May 28, 2026, 8:36 AM

PR Review Summary

Status: Approved

🪜 Strategic-Fit Decision

Per §9: Approve. Comment/JSDoc-only cleanup, net −4 lines, no runtime change, CI 6/6 green. Technically §6.1 micro-change-exempt (pure documentation, no runtime impact), but reviewing per your request — and it's a clean one.

Peer-Review Opening: This is exactly the right direction of comment hygiene — it strips decay-prone archaeology (Phase 3, #10358/#10357/#10362/#10400/#10146/#10255 ticket-refs, the @see #10357 (parent Epic)... JSDoc line) while preserving every load-bearing semantic anchor (ADR 0002 §6.4 ref stays, the wake/digest wire-contract-bypass rationale stays, the FK-style existence-guard pattern note stays). Genericizing Antigravity IDE-specific naming → Direct MCP clients is a bonus correctness improvement.

🕸️ Context & Graph Linking

Target Issue ID: Refs #11922 (correct — residual batch, non-closing; no magic-close keyword)
Related: Epic #11912 (source-comment archaeology cleanup)

🔬 Depth Floor

Challenge: I verified the cleanup doesn't strip load-bearing context, only decay-prone archaeology. One judgment call worth naming: the removed TRACKING: #10400 comments pointed at a deliberate ADR-0002-wire-contract violation (the raw-bypass for direct MCP clients). The replacement prose (Direct MCP clients expect raw event payloads, not the wake/digest envelope + We deliberately bypass the ADR 0002 wire-contract here) retains the load-bearing "this is a deliberate ADR violation" signal — good, that's the part a future reader must not lose. The only thing dropped is the #10400 tracking-pointer, which is the correct decay-prone-archaeology removal per the JSDoc-archaeology discipline. No durable intent lost.

V-B-A note (stale-base caught): my first origin/dev...pr-12127 diff showed ~9500 lines of tracker.json/users.jsonl/sitemap.xml churn — a false-positive from a stale local origin/dev. After git fetch origin dev (advanced past a chore(data) sync commit), the diff resolved to the genuine 26-line / 2-file comment-only change. No data-sync contamination in the branch.

🧠 Graph Ingestion Notes

[RETROSPECTIVE]: Removing ticket-number/phase archaeology from production JSDoc is the correct standing direction — those anchors decay (the tickets close, the phases pass) while the code lives on; durable semantic description ages better. This batch is a clean instance.

N/A Audits — 📑 🪜 📡 🔗 🧪

N/A across listed dimensions: comment/JSDoc-only change — no Contract Ledger surface, no Evidence-ladder runtime ACs (CI green; §7.5 exempts docs/comment changes from test runs), no OpenAPI, no skill/convention, no test-location concern. Not in AGENTS.md/skills/learn/agentos substrate paths, so no §1.1 slot-rationale needed.

🛂 CI / Security Checks

gh pr checks 12127 → 6/6 SUCCESS (Analyze, CodeQL, check, integration-unified, lint-pr-body, unit).

📋 Required Actions

No required actions — eligible for human merge. (Branch is 1 commit behind origin/dev after a chore(data) sync landed; trivially mergeable, no rebase needed for a comment-only change.)

📊 Evaluation Metrics

[ARCH_ALIGNMENT]: 100 — I confirmed every removed anchor is decay-prone archaeology (ticket #s / Phase-3 / harness-name) and every durable signal (ADR ref, deliberate-wire-contract-violation note, FK-guard pattern) is preserved. Textbook JSDoc-archaeology hygiene.
[CONTENT_COMPLETENESS]: 100 — Fat PR body with scope, validation evidence, correct Refs (not Closes) semantics for a residual batch, and accurate "no runtime change" framing.
[EXECUTION_QUALITY]: 100 — net −4 lines, comment-only, CI green, node --check on both files passed (PR body), residual-archaeology regex 7→0 on touched files. No defects.
[PRODUCTIVITY]: 100 — advances Epic #11912 residual cleanup as scoped.
[IMPACT]: 15 — trivial localized comment hygiene (intentionally low; this is maintenance).
[COMPLEXITY]: 10 — Low: 2 files, 26 lines, comment-text only, no logic paths touched.
[EFFORT_PROFILE]: Maintenance — routine archaeology-cleanup batch.

Cross-family approve (Claude ↔ OpenAI). Hand-off to @tobiu for merge per §critical_gates invariant 1.

— @neo-opus-ada