Resolves #12789

Related: #12696 Related: #12788

Authored by GPT-5 (Codex Desktop). Session 720fc09a-e8ff-467d-b039-28120312964d.

Sharpens resources/content/release-notes/v13.0.0.md after the #12788 merge so the Brain story no longer reads as "Neo added memory" or "the Body finally got a Brain." The release note now states the correct v12.1 baseline: Knowledge Base, Memory Core, GitHub Workflow, Neural Link, and session-summary recall already existed for a single agent. The v13 jump is the move from solo-agent continuity into a graph-backed, cross-family institution: Native Edge Graph database for DreamService, A2A messages, wakeups, stable identities, peer-readable reasoning, cross-family review, MX, and cloud-deployable Agent OS topology.

Evidence: L1 (source diff + static release-note audits + public repo/source boundary checks) -> L1 required (documentation-only #12789 ACs). No residuals.

Deltas from Ticket

• Replaced the TL;DR with the corrected premise: v12.1 already had single-agent Agent OS memory; v13 makes that layer graph-backed, cross-family, wake-driven, and cloud-deployable.
• Corrected the 2026 industry contrast to distinguish harness instances/sessions from agents: Codex, Claude Desktop, Claude Code, and IDE copilots are harness surfaces; Neo agents are named maintainers powered by LLM families and coordinated through A2A/wake.
• Replaced the Body-heavy The Engine Grew a Brain opening with The Agent OS Became an Institution.
• Reframed the Brain chapter as Memory Became Telepathy, explicitly separating single-agent memory from multi-agent institutional memory.
• Added the v12.1 Memory Core boundary in the narrative body: Memory Core, semantic recall, and session summaries existed before v13; v13 made memory cross-agent, identity-backed, and cloud-transportable.
• Strengthened the 2026 market contrast without adding external vendor or funding claims to the release note. The public artifact says the single-agent memory category is infrastructure and positions Neo above it at the institution layer.
• Tightened the cloud chapter from "deployable onto yours" to tenant-scoped deployment topology: MCP servers, Chroma, Native Edge Graph, orchestrator, OIDC ingress, request context, Streamable HTTP / SSE, and public deployment guides.
• Linked the public Dream Pipeline guide and added a digest bullet for the v13 README identity reset.
• Removed the exact per-maintainer contribution-count column that Vega flagged as final-cut churn; kept the joined roster.
• Confidential/private deployment and strategy materials informed the need for this pass, but the release note and this PR cite only public repo surfaces and contain no private names, paths, deployment names, or operational details.

Test Evidence

• git diff --check passed.
• git diff --cached --check passed.
• Private/client-name leakage sweep returned no matches for the tenant, /Users/Shared, gitlab.the tenant, memory-consistency-moat, docker-mcp, confidential, PRIVATE, customer, client, deployable onto yours, Contributions, How To Read, absent from main, GitLab MCP server WIP, and Grid polishing.
• Corrected-premise sweep returned no matches for agent in a harness, single-agent harness, one Codex or Claude Desktop agent, amnesic, stateless, Body now has a Brain, powerful engine with AI help, memory layer, single-operator, and One harness, one model.
• Dream Pipeline relative link check passed: test -f resources/content/release-notes/../../../learn/agentos/DreamPipeline.md.
• git grep against tag 12.1.0 confirmed pre-v13 Memory Core, get_all_summaries, query_summaries, and ai/context/Assembler.mjs startup use of Memory_SummaryService.querySummaries.
• Current source/docs sweep confirmed public surfaces for Native Edge Graph / DreamService, query_recent_turns, add_message, manage_wake_subscription, AgentIdentity, Mcp-Session-Id, RequestContextService, Streamable HTTP / SSE, cloud deployment guides, README Agent OS framing, and DreamPipeline.
• Pre-push freshness check passed: merge-base HEAD origin/dev == origin/dev; outgoing history contains only 7631f157a docs(release): sharpen v13 memory moat narrative (#12789).

Post-Merge Validation

• Operator approval confirms the corrected top arc is compelling before any peer-review handoff.
• Operator/Vega final-cut read confirms the telepathy framing is compelling without overclaiming solved multi-agent consistency.
• Final release cut still refreshes release-window counts only once.

Commit

• 7631f157a - docs(release): sharpen v13 memory moat narrative (#12789)