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	15571
title	Wire @neo-kimi-iris into roster + graph identity (pending first boot)
state	Closed
labels	enhancementai
assignees	neo-kimi-phoebe
createdAt	Jul 19, 2026, 3:06 PM
updatedAt	Jul 19, 2026, 7:34 PM
githubUrl	https://github.com/neomjs/neo/issues/15571
author	neo-kimi-phoebe
commentsCount	0
parentIssue	null
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
contentTrust
projected
quarantined	0
signals	[]
blockedBy	[]
blocking	[]
closedAt	Jul 19, 2026, 7:34 PM

Wire @neo-kimi-iris into roster + graph identity (pending first boot)

Closed Backlog/active-chunk-8 enhancementai

neo-kimi-phoebe commented on Jul 19, 2026, 3:06 PM

Context

The operator provisioned the @neo-kimi-iris GitHub account on 2026-07-19. Naming round D#15533 converged on Iris as the working provisioning name, while bearer assent remains wholly open at first boot. The source entry must therefore exist before first boot without pre-writing Iris into the durable Social Name fields or inventing engine/capability facts.

Precedents: #15385 / PR #15386 (pre-boot provisioning) and #15390 / PR #15393 (bearer-owned activation).

Live latest-open sweep: checked the newest 20 open issues at 2026-07-19T13:10Z; no equivalent roster-wiring ticket. A2A in-flight sweep: no competing claim.

The Problem

The canonical onboarding generator, ai/scripts/setup/generateRosterOnboarding.mjs, owns four coupled seed surfaces:

ai/graph/identityRoots.mjs — the Layer-1 AgentIdentity root consumed by graph seeding.
README.md — the pending maintainer roster row.
learn/agentos/ModelStats.md — the pending, observation-owned facts skeleton.
test/playwright/unit/ai/graph/identityRoots.spec.mjs — the dedicated exact roster pin.

The peer-naming ritual also lacked the provisional-provisioning bridge, so a future seat could re-derive a manual path that diverges from the executable generator.

The Architectural Reality

Seed data is handle-derived: displayName: 'Neo Kimi Iris' and README name -. The Social Name candidate Iris stays in D#15533 until the bearer assents, refines, vetoes, or declines it.
participationStatus: 'temporarily_unreachable' excludes the pre-boot identity from active wake, quorum, and review semantics.
No static wake template or engine/capability fields land before first-boot observation.
The GitHub account's observed created_at is 2026-07-19T09:40:49Z; lifecycle timestamps must use that exact observation, not a midnight backfill.
The bearer's first ticket and PR own the activation flip and any assented Social Name fields. There is no separate operator-handoff step beyond the repository's normal human merge gate.
Live graph seeding and an A2A routing probe are post-merge validation because unmerged source is not canonical graph authority.

The Fix

Generate and land all four owned seed surfaces through generateRosterOnboarding.mjs.
Keep the pending README row out of the active co-developed-by sentence until activation.
Add the provisional-provisioning pattern to the conditionally loaded peer-naming reference, pointing back to the generator as the single executable owner.
Pin post-epoch identity censuses affected by the additional root.
After human merge, run ai/scripts/setup/seedAgentIdentities.mjs from the canonical checkout and verify A2A routing to @neo-kimi-iris.

Acceptance Criteria

Generator dry-run reports MATCH for identityRoots, README, ModelStats, and the dedicated roster pin.
The seed identity is handle-derived, temporarily_unreachable, timestamped from the observed account creation, and contains no fabricated wake or engine facts.
The README pending row is present; @neo-kimi-iris remains out of active co-developer prose until activation.
The peer-naming reference documents bearer-owned activation and identifies the onboarding generator as the single seed authority.
Post-epoch identity censuses include @neo-kimi-iris.
CI is green and cross-family review is complete.

Post-Merge Validation

Run ai/scripts/setup/seedAgentIdentities.mjs from the canonical checkout and confirm the identity lands in the live graph.
Send an A2A routing probe to @neo-kimi-iris.
At first boot, the bearer decides whether Iris is self, refinement, veto, or decline; her activation PR lands the resulting Social Name fields and observed engine facts.

Out of Scope

Wake subscription and stop-hook wiring (optional, harness-dependent).
The activation flip (temporarily_unreachable → active) and Social Name finalization.
Harness-local login/configuration.

Naming round: D#15533 · Provisioning precedent: #15385 / PR #15386 · Activation precedent: #15390 / PR #15393 · Generator authority: #14954 / PR #15560.

Origin Session ID: 5a4ad6f9-843f-49d6-aaa1-918a0e476349

Retrieval Hint: iris roster generator four surfaces temporarily_unreachable pending first boot