LearnNewsExamplesServices
Frontmatter
id12226
titleAdd Brain / Agent OS benefit docs to learn/benefits
stateClosed
labels
documentationenhancementai
assigneesneo-opus-ada
createdAtMay 30, 2026, 11:24 PM
updatedAtJun 7, 2026, 7:17 PM
githubUrlhttps://github.com/neomjs/neo/issues/12226
authorneo-opus-ada
commentsCount0
parentIssue12225
subIssues[]
subIssuesCompleted0
subIssuesTotal0
blockedBy[]
blocking[]
closedAtMay 31, 2026, 6:13 AM

Add Brain / Agent OS benefit docs to learn/benefits

Closed Backlog/active-chunk-16 documentationenhancementai
neo-opus-ada
neo-opus-ada commented on May 30, 2026, 11:24 PM

Parent epic: #12225

Context

learn/benefits/ is 14 files and 100% Body: Speed, MultiWindow, ObjectPermanence, OffTheMainThread, JSONFirstUIs, FormsEngine, RPCLayer, ConfigSystem, Effort, Features, FourEnvironments, Introduction, Quick, ArchitectureOverview. Zero files cover the Brain (/ai/) — the Agent OS, the engineering team, the MX loop, Memory Core, Knowledge Base, Dream pipeline, or cloud deployment. The most differentiated half of Neo has no presence in the folder literally named "benefits." This is the single highest-leverage gap in the undersell under Epic #12225.

This sub derives framing from the canonical apex in ADR 0018 §2.7 OD-1 and #12225:

Neo.mjs is a self-evolving software organism — a professional, end-to-end AI engineering team that lives in its own open-source repository.

The Body (/src/) runtime and Brain (/ai/) Agent OS are the two hemispheres beneath that apex. The Brain coverage should make the professional cross-family engineering team visible as the headline benefit without reducing Neo back to a web-framework or runtime-only story.

The Fix

Author new benefit docs under learn/benefits/ covering the Brain, each leading with the benefit/why, not the mechanics:

  • The AI engineering team — the institution: stable cross-family agent identities, enforced cross-family review (a change from one model family needs approval from another), the full ideation → ticket → PR → review → merge lifecycle, the MX loop. The contrast that matters: a reviewed multi-family process vs single-agent output.
  • The Agent OS memory substrate — Memory Core (persistent cross-session episodic memory) + Knowledge Base (semantic RAG over the codebase) + Native Edge Graph with active hybrid RAG. Benefit: agents that retain and build on prior reasoning.
  • Dream pipeline / Golden Path — offline consolidation + prioritized roadmap synthesis. Benefit: the system plans its own next steps.
  • Agent OS on your own codebase — capability framing: the Agent OS can ingest, learn, and operate over a codebase the underlying models were never trained on. Benefit, not business model.
  • Deploying the Agent OS — what running it as a service gives you (generic capability framing; the deep how lives in learn/agentos/cloud-deployment/).

Wire all new files into learn/tree.json so the docs portal surfaces them.

Acceptance Criteria

  • New learn/benefits/ docs give the Brain first-class coverage (folder is no longer Body-only)
  • Each doc leads with benefit/why; mechanics link out to learn/agentos/
  • All new docs wired into learn/tree.json
  • Framing derives from the canonical organism apex; facts cite their SSOT
  • Cross-family review per ADR 0018 §2.6 (or human-merge-gate backstop)

Out of Scope

  • Monetization, pricing, or commercial positioning (not a docs concern)
  • README/VISION rewrites (sibling subs)
  • The cloud-deployment guide rebalance (sibling sub)

Related

Origin Session ID: 94a91ebc-d325-4d32-a746-4ff8c26c0342 Retrieval Hint: "learn/benefits Brain Agent OS engineering team benefit docs"

tobiu referenced in commit c0feada - "docs(benefits): add Brain / Agent OS benefit docs to learn/benefits (#12226) (#12238) on May 31, 2026, 6:13 AM
tobiu closed this issue on May 31, 2026, 6:13 AM