Close FABRIC-2.md, open FABRIC-3.md; audit Stadium dispatch stubs; new no-stubs rule

FABRIC-2.md closed archival at ~4,400 lines / 51 open items, same reasoning
FABRIC.md itself was closed for at 7,595 lines -- continuing to append made
still-open work hard to find. All 51 open items carried forward into
FABRIC-3.md's new Section A, verified complete via programmatic diff against
the source (49 unique + 2 confirmed pure duplicates from FABRIC-2.md's own
F.3 cross-reference section, not dropped content). .claude/CLAUDE.md's
pointer note updated to name FABRIC-3.md as current.

New Section B: full audit of stadium_dispatch()'s four behaviour stubs
(MIGRATE/DELIVER/EXPIRE/COOL), triggered by investigating "words/VMs/blocks/
messages should all be on the same engine". Found the picture is more
nuanced than "everything's a stub" -- words are already fully live via a
separate bespoke mechanism (stadium_word_dispatch(), wired into vm_core.c's
real word-execution path, item 4.1), contradicting ONTOLOGY.md's stale
"not yet migrated" claim (flagged for a follow-up fix). MIGRATE (blocks)
and VM-COOL are genuinely stub with zero consumers; DELIVER (Hermes) and
EXPIRE (ACL) are substantial, undecided subsystem integrations FABRIC.md
itself already flagged as open, not touched here.

New standing rule, saved as memory feedback_no_stubs_or_todos.md: stub
implementations and TODO placeholders are never acceptable in this
workflow, in any language, ever -- triggered by finding stadium_dispatch()'s
stub handlers during this investigation.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CXjAPTEKrgY2Mrk25KoLDn

.claude/CLAUDE.md

@@ -10,11 +10,12 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
 
 > **Superseded subsystem docs (Captain Bob, 2026-08-15):** `.claude/TRIPOD.md`,
 > `.claude/HERMES.md`, `.claude/ARTEMIS.md`, and `.claude/CONSOLE.md` are all superseded —
-> `FABRIC.md` (design history) and `FABRIC-2.md` (current/living, read this one first) are
-> the sole authoritative source for Tripod/Hermes/Artemis/Console work now. The four
-> subsystem docs remain in the repo as historical record only; each carries its own
-> superseded-header pointing here. Do not treat them as current, do not read them for design
-> authority, and do not cite them in place of `FABRIC.md`/`FABRIC-2.md`.
+> `FABRIC.md` and `FABRIC-2.md` (both design history/archival as of 2026-08-25) and
+> `FABRIC-3.md` (current/living, read this one first) are the sole authoritative source for
+> Tripod/Hermes/Artemis/Console work now. The four subsystem docs remain in the repo as
+> historical record only; each carries its own superseded-header pointing here. Do not treat
+> them as current, do not read them for design authority, and do not cite them in place of
+> `FABRIC.md`/`FABRIC-2.md`/`FABRIC-3.md`.
 
 > **Scope:** This repo is LithosAnanke — the bare-metal UEFI kernel that boots StarForth
 > directly on hardware. StarForth (the hosted FORTH-79 VM) has its own separate repository