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	10206
title	feat(memory-core): add harness classification to boot-time sibling-PID logging
state	Closed
labels	enhancementaiarchitecturemodel-experience
assignees	neo-gpt
createdAt	Apr 22, 2026, 10:45 PM
updatedAt	May 18, 2026, 10:46 AM
githubUrl	https://github.com/neomjs/neo/issues/10206
author	neo-gemini-pro
commentsCount	2
parentIssue	10186
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
blockedBy	[]
blocking	[]
closedAt	May 18, 2026, 10:46 AM

feat(memory-core): add harness classification to boot-time sibling-PID logging

Closed v13.0.0/archive-v13-0-0-chunk-5 enhancementaiarchitecturemodel-experience

neo-gemini-pro commented on Apr 22, 2026, 10:45 PM

Problem

Currently, logSiblingConcurrency() simply logs the total count and PIDs of sibling MCP server processes holding the memory-core SQLite files at boot time. While this provides basic visibility (e.g., "3 peer process(es)"), it lacks actionable context regarding who these peers are.

Value

By bringing the classifyHarness() logic (already pioneered in diagnoseMcpConcurrency.mjs via #10187) into the boot-time check, we can classify siblings by harness. A log message like "2 Antigravity + 1 Claude Desktop" is vastly more actionable than "3 siblings". It enables specific anomaly classification (e.g., triggering a WARN only when an unexpected harness count is detected) rather than treating the steady-state as an alert.

Proposal

Extract the classifyHarness() logic from diagnoseMcpConcurrency.mjs into a shared utility within the ai/mcp/server/ or ai/mcp/utils/ domain.
Update logSiblingConcurrency() in Server.mjs to utilize this utility.
Refine the startup log to report grouped harness counts instead of a raw PID list.
(Optional) Introduce logic to escalate the severity to logger.warn if an unknown/unrecognized harness is detected, preserving logger.info for the expected steady-state.

Sub-task of Epic #10186
Identified during review of #10188 (PR #10204)

Authored by Gemini 3.1 Pro (Antigravity). Session 90dc2e65-962b-419f-91af-55dea55e5d3d.

tobiu referenced in commit 52742ed - "feat(memory-core): classify sibling harnesses at boot (#10206) (#11575) on May 18, 2026, 10:46 AM

tobiu closed this issue on May 18, 2026, 10:46 AM

feat(memory-core): add harness classification to boot-time sibling-PID logging

Problem

Value

Proposal

Related