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.

Neo.mjs - Self-Evolving Software Organism

Frontmatter

title	>-
author	tobiu
state	Merged
createdAt	Apr 13, 2026, 1:57 PM
updatedAt	Apr 13, 2026, 1:59 PM
closedAt	Apr 13, 2026, 1:59 PM
mergedAt	Apr 13, 2026, 1:59 PM
branches	`dev` ← `agent/9965-map-reduce-session`
url	https://github.com/neomjs/neo/pull/9966

Merged

tobiu commented on Apr 13, 2026, 1:57 PM

Resolves #9965. Refactors SessionService into a lossless Map-Reduce chunking pipeline. Drops the aggressive historical truncation block installed in #9921, securing massive RLAIF extraction loops by preserving their full contextual path. Features complete Playwright unit test mapping emulation.

tobiu commented on Apr 13, 2026, 1:58 PM

Input from gemini-exp-1206 (Antigravity):

✦ # PR Review Summary

Status: Approved

Awesome refactoring! This guarantees we no longer blindly chop off episodic memory when saving large tasks. Offloading the context parsing to a Map-Reduce loop directly within SessionService secures the long-term reliability of our internal RLAIF structure.

📊 Evaluation Metrics

[ARCH_ALIGNMENT]: 100 - Solved a complex token limit issue perfectly within the existing abstractions (the internal this.model.generateContent API layer).

[CONTENT_COMPLETENESS]: 95 - The implementation handles the sequential splitting gracefully and adds important console traces for when Map-Reduce fires.

[EXECUTION_QUALITY]: 100 - Playwright tests specifically model the Map-Reduce loop to prevent regressions, isolating mock detection logic elegantly.

[PRODUCTIVITY]: 100 - Resolved P0 Ticket #9965.

[IMPACT]: 90 - Protects against systemic data loss and solidifies the robustness of graph extraction.

[COMPLEXITY]: 50 - Required an asynchronous map stage and array orchestration within an already complex string parsing method.

[EFFORT_PROFILE]: Heavy Lift - Resolves a structurally compromising issue dynamically.

🕸️ Context & Graph Linking

Target Epic / Issue ID: Resolves #9965

Related Graph Nodes: SessionService

🧠 Graph Ingestion Notes

[KB_GAP]: N/A

[TOOLING_GAP]: N/A

[RETROSPECTIVE]: Token exhaustion is best mitigated actively inside the infrastructure layer through lossless chunking (Map-Reduce) instead of aggressive historical truncation. This preserves the internal chronological sequence of an agent's hypotheses and actions seamlessly across local API endpoints.

📋 Required Actions

To proceed with merging, please address the following:

Merge PR #9966 to dev using Squash Merge.

Looks ready to go!