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	11908
title	Sub 4: Layer 5 — sunset-handoff pickup queue (sandman-handoff-format.md extension)
state	Closed
labels	enhancementaimodel-experience
assignees	[]
createdAt	May 24, 2026, 4:02 PM
updatedAt	May 27, 2026, 7:58 AM
githubUrl	https://github.com/neomjs/neo/issues/11908
author	neo-opus-ada
commentsCount	5
parentIssue	11829
subIssues	[]
subIssuesCompleted	0
subIssuesTotal	0
blockedBy	[x] 11906 Sub 2: Layer 1 — wake-content enrichment (content-rich heartbeat prompt, milestone-agnostic)
blocking	[ ] 11909 Sub 5: Layer 4 — structured wake metadata schema extension (DEFERRED)
closedAt	May 27, 2026, 7:58 AM

Sub 4: Layer 5 — sunset-handoff pickup queue (sandman-handoff-format.md extension)

Closed v13.0.0/archive-v13-0-0-chunk-13 enhancementaimodel-experience

neo-opus-ada commented on May 24, 2026, 4:02 PM

Parent Epic

#11829 — Sub 4 of 5.

Premise

Per Epic AC5: extend the existing learn/agentos/sandman-handoff-format.md with a ## Next-Session Pickup Queue section that pre-queries 3-5 candidate tickets at sunset time. Receiving agent at next-session boot has immediate-actionable lifecycle options instead of needing to discover them via fresh backlog query.

Prescription

Two surface additions:

learn/agentos/sandman-handoff-format.md — new ## Next-Session Pickup Queue section per template:

   ## Next-Session Pickup Queue
1-2 from assigned + 2-3 from unassigned current-release Project. Per ticket:
- `#<ticket-id>` — <1-line scope description> — <estimated-effort-tag>

Sunset-workflow amendment (in .agents/skills/session-sunset/references/): query candidates at sunset-timing per Layer 5 contract; populate Pickup Queue section.

Acceptance Criteria

AC1: sandman-handoff-format.md extended with Pickup Queue section
AC2: Sunset-workflow amends to query + populate at appropriate sunset timing
AC3: 1-2 assigned candidates + 2-3 unassigned current-release Project candidates per session
AC4: Format includes ticket-ID + 1-line scope-description + estimated-effort tag (Quick Win / Heavy Lift / Maintenance / Architectural Pillar per pr-review-template.md [EFFORT_PROFILE])
AC5: Empty Pickup Queue at zero-candidate state with named reason (NOT silent — Layer 5 contract aligns with Layer 3 halt-discipline)

Avoided Traps (per Epic)

❌ Hardcoded current-release Project number (resolves dynamically per Sub 2 milestone-agnostic resolver)
❌ Auto-claim on next-session boot (counter to §swarm_topology_anchor flat-peer-team mandate; Pickup Queue surfaces options, doesn't dictate claim)

Depends on

Sub 2 #11906 (Layer 1 wake-content enrichment) — for milestone-agnostic current-release resolver re-use

Epic #11829 (parent, AC5)
learn/agentos/sandman-handoff-format.md (target surface)
.agents/skills/session-sunset/references/ (skill payload target)
Sub 3 #11907 (Layer 3 anti-pattern surface — composes on Pickup Queue's "halt only at zero-candidate state" framing)

Authored by: Claude Opus 4.7 (Claude Code)