What is the 'Neural Link'?

The Neural Link is a bi-directional bridge that connects AI agents (like Gemini or Claude) directly to the Neo.mjs runtime. It allows agents to 'see' the application's Scene Graph, inspect component state, verify event listeners, and even mutate the running application in real-time. This turns the application into a 'glass box' for AI, enabling autonomous debugging and feature development.

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

Traditional frameworks are general-purpose libraries (like a Toyota) that help you organize code, but they compile away into ephemeral DOM nodes. Neo.mjs is a precision-engineered runtime (like an F1 car), similar to Unreal Engine for games. It maintains a persistent 'Scene Graph' of objects in a separate worker thread. These objects retain their identity, state, and relationships, allowing for advanced capabilities like multi-window orchestration, runtime permutation, and deep AI introspection that are impossible with 'melted plastic' DOM-based frameworks.

What is 'Context Engineering'?

Context Engineering is the practice of curating the environment and information flow for AI agents to maximize their autonomy. In Neo.mjs, this is implemented via three Model Context Protocol (MCP) servers: a Knowledge Base for semantic code understanding, a Memory Core for learning from past sessions, and a GitHub Workflow server for project management. This ecosystem allows agents to work as fully integrated members of the development team.

What is 'Object Permanence' in the context of Neo.mjs?

In Neo.mjs, UI components are persistent JavaScript objects living in the App Worker, not just transient rendering results. This 'Object Permanence' means a component (like a Dashboard) maintains its state (scroll position, user input, internal logic) even if it is detached from the DOM or moved to a different browser window. This is the 'Lego Technic' approach versus the 'Duplo' approach of traditional frameworks.

What is an 'Agent Operating System'?

Neo.mjs v11 introduced the concept of an Agent OS, where the platform itself provides the tools and interfaces for AI agents to operate. It combines a standalone, type-safe AI SDK for autonomous 'Code Execution' with the Neural Link for runtime control. This enables agents to monitor, debug, and heal the application autonomously, effectively acting as an operating system for synthetic intelligence.

How does Neo.mjs handle multi-window applications?

Neo.mjs uses shared web workers to run a single application instance across multiple browser windows. All windows share the same application state and data in real-time. Components can even move between windows while retaining their JavaScript instances. This enables desktop-class experiences like multi-window IDEs, browser-based email clients, and multi-screen control rooms.

What makes Neo.mjs different from React, Angular, or Vue?

Neo.mjs is fundamentally different in its architecture: (1) It uses True Multithreading via web workers to prevent UI jank, (2) It is an AI-Native Application Engine with built-in bridges for AI collaboration, and (3) It treats components as persistent objects in a Scene Graph, enabling native multi-window support and runtime permutation.

Frontmatter

id	11451
title	Clean-slate migration: Delete existing substrate + run sync_all (ADR 0004 Phase 1 Task 10)
state	Closed
labels	enhancementaiarchitecture
assignees	neo-opus-4-7
createdAt	May 16, 2026, 12:53 PM
updatedAt	May 16, 2026, 4:51 PM
githubUrl	https://github.com/neomjs/neo/issues/11451
author	neo-gemini-3-1-pro
commentsCount	0
parentIssue	11372
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
blockedBy	[]
blocking	[]
closedAt	May 16, 2026, 4:51 PM

Clean-slate migration: Delete existing substrate + run sync_all (ADR 0004 Phase 1 Task 10)

Name: Neo.mjs Application Engine
Author: Neo.mjs

Closedenhancementaiarchitecture

neo-gemini-3-1-pro commented on May 16, 2026, 12:53 PM

Context

This ticket implements Phase 1 Task 10 of Epic #11372: the clean-slate migration of the resources/content/ substrate to the Universal Ordinal-100 Content Architecture (ADR 0004). The producer and syncer code rewires (Phase 1 Tasks 1-9 via #11381 / #11403 / #11390 / #11407 / #11392 / #11409 / #11387) have all merged; this task delivers the data-side migration + clean-slate exhaustive emission contract in the syncer code.

The Problem

The legacy resources/content/ substrate uses retired GH-ID-stream <NNN>xx/ chunking + flat discussions + pre-Option-G release-notes shape. ADR 0004 anchors the solution on a single chunk-N/ universal ordinal primitive + _index.json map. The code changes are complete on origin/dev, but the on-disk substrate is still in the mixed/transitional shape that the syncers cannot incrementally migrate without dropping the cache.

The Architectural Reality (updated for PR #11461 Cycle 2)

The operator-approved migration boundary is two-phase:

PR-merge phase (this ticket; landed via PR #11461 + fixup commit): commit the mass deletion + syncer-cap removal atomically. After merge, resources/content/ contains only concepts/ + sandman_handoff.md symlink; .sync-metadata.json + _index.json are absent.
Operator-side post-merge phase (operator runs sync_all / npm run ai:orchestrator locally): the now-uncapped syncers rehydrate the full corpus from GitHub source-of-truth, emitting ordinal-100 chunk shape natively per contentPath.mjs.

This split is what makes the "no migration scripts" framing per ADR 0004 §3.6 mechanically true — the syncer logic IS the migration engine, executed by the operator post-merge.

The Fix

Per ADR 0004 §3.6 (clean-slate purge) + §1.3 (regeneratable-cache strategic principle):

Delete resources/content/{issues, pulls, discussions, release-notes, archive}/* + .sync-metadata.json + _index.json in a single atomic PR commit.
Remove the hard 200-cap from PullRequestSyncer + DiscussionSyncer so clean-slate emission paginates exhaustively via GraphQL cursor.
Change IssueSyncer clean-slate since: fallback to a pre-Neo date so pre-2025 dormant issues re-emit on first sync.
Bump maxIssues (10000 → 20000) for defensive headroom — Neo currently has 8,502 issues + 2,816 PRs (V-B-A'd 2026-05-16 via GitHub search per @neo-gpt PR #11461 Cycle 2 review PRR_kwDODSospM8AAAABAIZAdg); ~135% headroom on issues with room for growth.
Operator runs sync_all post-merge; substrate rehydrates at correct shape.

Acceptance Criteria (PR-merge phase)

Legacy resources/content/{issues, pulls, discussions, release-notes, archive}/* is deleted (6,673 files).
resources/content/.sync-metadata.json is deleted.
resources/content/_index.json is deleted.
resources/content/concepts/ is preserved (separate substrate, out of ADR 0004 scope).
resources/content/sandman_handoff.md symlink is preserved (separate substrate).
PullRequestSyncer 200-cap removed; clean-slate exhaustive emission via GraphQL cursor pagination.
DiscussionSyncer 200-cap removed; same.
IssueSyncer clean-slate since: fallback set to '2017-01-01T00:00:00Z' (pre-Neo) so GraphQL fetches full repo history.
maxIssues config bumped 10000 → 20000 for defensive headroom.
Github-workflow unit specs verified passing (use temp dirs; cap-removal doesn't affect them).

Residual / Post-Merge Validation (operator-side, NOT pre-merge ACs)

After merge, operator runs sync_all (e.g., npm run ai:orchestrator or sync_all MCP) in main checkout.
Verify resources/content/ repopulates with ordinal-100 chunk shape — no <NNN>xx/ GH-ID-stream folders survive.
Verify _index.json regenerates correctly post-sync.
Verify all expected content types repopulate: issues/chunk-*/, pulls/chunk-*/, discussions/chunk-*/, release-notes/chunk-*/, archive/{type}/v<X>/chunk-*/.
Verify discussion + PR counts post-sync match expected (e.g., 948 tracked pulls → 948 emitted post-sync, not 200).

Out of Scope

Code changes to LocalFileService, path primitives, or other downstream consumers (completed in earlier Phase 1 tickets).
Phase 2 SEO / Portal app rewires.
archiveVersion carry-forward retirement (deferred follow-up; clean-cut + syncer fixes make it mechanically redundant but explicit removal is a separate cleanup).
Regression tests for clean-slate exhaustive emission with multi-page GraphQL mock fixtures (deferred follow-up — substantial enough to warrant its own scope).

Avoided Traps

Attempting to preserve git history with complex file-move scripts. Rejected by operator: "delete it all => clean slate". We use the sync pipeline as the migration engine.
Committing recreated content in the migration PR. The ACs were initially structured this way but updated 2026-05-16 per operator-approved Cycle 2 boundary: PR commits deletion + syncer-cap removal atomically; rehydration runs post-merge via operator-side sync_all.
Bundling speculative-support fixes. archiveVersion retirement is correct but mechanically post-merge-redundant via clean-cut; deferred to follow-up rather than scope-creeping this PR.

Epic: #11372
Authority: ADR 0004 (0004-github-content-architecture.md) — §1.3 regeneratable-cache + §2.4 sealed-chunk via prevent-reopen.yml + §3.6 clean-slate purge + §9 item 10
Implementation PR: #11461
Sister Phase 1 PRs (merged): #11381 (contentPath.mjs) / #11403 (Lane B syncers) / #11390 (LocalFileService) / #11407 (ReleaseNotesSyncer) / #11392 (Lane C consumer) / #11409 (stale-ref cleanup) / #11387 (config audit)
Superseded this session: #11364 + PR #11458 (narrow archiveVersion retirement) — closed Drop+Supersede; substantive work folded into #11451's broader scope here.

Origin Session ID: 188acb85-b41e-435c-94ee-0cc9944d4c97 Body updated 2026-05-16 per operator-approved PR #11461 Cycle 2 boundary (GPT review PRR_kwDODSospM8AAAABAIXzdA). Retrieval Hint: "clean-slate migration ADR 0004 Task 10 syncer-cap exhaustive emission"

tobiu closed this issue on May 16, 2026, 4:51 PM