FABRIC-3.md: close the live-MIGRATE verification gap; report a USE/RUN shadowing bug

Confirmed the real MIGRATE dispatch fires live: a temporary boot probe
(STADIUM-ADMIT + STADIUM-EVICT against the live Artemis VM, inserted
into the existing 4.6 self-test and reverted immediately after capture)
produced "Stadium: dispatch cell=63257 behaviour=MIGRATE lbn=100" with
blk_flush(100) firing and zero error -- closes the honest gap left open
in the previous commit. Interactive flooding alone couldn't reach this:
Hera's reservoir sits at the Q48_ONE/3 floor from boot self-tests, so
block-touch candidates pull 0 heat and can never out-density an
existing resident, a pre-existing reservoir-floor/eviction interaction
unrelated to this pass.

Also reports (not fixes, per CLAUDE.md) a real dictionary-shadowing bug
found while chasing this: capsules/lib.4th:13 redefines USE as EXEC,
shadowing the C primitive mama_word_use() (REPL VM-redirect) with an
unrelated capsule-loading meaning -- same for RUN at lib.4th:14. Same
bug class as the K-PUSH dictionary-shadowing issue.

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

FABRIC-3.md

@@ -317,18 +317,37 @@ other three become honest, explicit punch-list items instead of being touched sp
   identically across all three; `BLOCK`/`BUFFER` touches exercised live from the REPL on amd64 and
   riscv64 with no crash; a 22,000-distinct-block flood loop (amd64, artificially shrunk to a
   20,971-cell Stadium via a one-off smaller `-m` to make quota pressure reachable) ran clean under
-  heavy admission-path load with no corruption. **Honest gap, not silently claimed:** a live
+  heavy admission-path load with no corruption. **Live `MIGRATE` fire confirmed (2026-08-26):**
-  `MIGRATE lbn=...` console print was NOT directly observed firing in this session. Root cause
+  interactive flooding alone never triggered it — Hera's reservoir was already sitting exactly at
-  traced, not a defect in this work: Hera's reservoir was already sitting exactly at the
+  the `Q48_ONE / 3` floor from the boot-time self-tests, so every block-touch candidate pulled 0
-  `Q48_ONE / 3` floor from the boot-time self-tests before any manual testing began, so every
+  heat, and a 0-heat candidate can never be *strictly denser* than an existing resident, so
-  block-touch candidate pulled 0 heat — and a 0-heat candidate can never be *strictly denser* than
+  `stadium_admit()`'s eviction fallback correctly refuses rather than evicts once the free list is
-  an existing resident, so `stadium_admit()`'s eviction fallback correctly refuses rather than
+  exhausted (a pre-existing reservoir-floor/density-eviction interaction, applies equally to word
-  evicts once the free list is exhausted. This is a pre-existing interaction between the reservoir
+  patrons, not introduced by this pass). Closed deterministically instead with a temporary boot
-  floor and density-based eviction (applies equally to word patrons, not something this pass
+  probe (`kernel_main.c`, inserted into the existing Artemis 4.6 self-test block, reverted
-  introduced), not something wrong with the new code. Follow-up: force a live `MIGRATE` fire from
+  immediately after capture — no code left behind): `100 65536 0 STADIUM-ADMIT ... STADIUM-EVICT`
-  inside a child VM's own REPL context (where the `STADIUM-ADMIT`/`STADIUM-EVICT` diagnostic words
+  against the live Artemis VM. Captured live: `Stadium: dispatch cell=63257 behaviour=MIGRATE
-  are registered — attempted via `USE` this session, which failed with "EXEC: failed", not yet
+  lbn=100`, followed by `DBG err after MIGRATE probe=0` — `blk_flush(100)` fired for real, with
-  root-caused) to directly and deterministically exercise the eviction path with nonzero heat.
+  the correct LBN threaded through from the departing patron's `identity` field exactly as
+  designed. (Artemis's own reservoir went briefly out of Q48_ONE-balance during the probe — raw
+  `STADIUM-ADMIT` doesn't debit the reservoir on its own, by its own doc, so a heat value handed
+  to it directly is invented, not pulled; harmless here since Artemis is killed and her whole
+  economy discarded immediately after, and Hera's own conservation was independently confirmed
+  back at the normal 43691/21845/65536 baseline afterward.)
+- [x] **Bug found and reported, not fixed (2026-08-26): `capsules/lib.4th:13` shadows the C
+  primitive `USE`.** While chasing the live-MIGRATE test above, `S" Artemis" USE` (meant to
+  redirect the REPL into Artemis's own vocabulary, `mama_word_use()`) instead printed `EXEC:
+  failed: Artemis`. Root cause: `capsules/lib.4th:13` defines `: USE ( addr u -- ) EXEC ;` — a
+  FORTH word with the same name but a completely different meaning ("load/exec a capsule"),
+  which shadows the C-registered `USE` in dictionary search order since `lib.4th` loads after
+  primitive registration. `lib.4th:14` does the identical thing to `RUN` (`: RUN ( addr u -- )
+  EXEC ;`), which CLAUDE.md also names as an untouchable C primitive ("BIRTH, RUN, USE are
+  primitives — registered in C exactly like DUP, BYE, EXEC"). Same bug class as the K-PUSH
+  dictionary-shadowing issue (`docs/working/architecture/K-PUSH-DICTIONARY-SHADOWING-BUG-
+  20260704.md`). Per CLAUDE.md's rule against unrequested fixes, this is reported here, not
+  patched — `lib.4th`'s `USE`/`RUN` were presumably meant as private helpers for
+  `doe-campaign.4th`-style capsules and never intended to collide with the REPL primitives of
+  the same name.
 - [ ] Scope VM-`COOL`: confirm whether `capsule_vm_kill()`/Tripod's existing VM lifecycle has
   any real hook point equivalent to `stadium_evict()`, or needs the same kind of admission
   work `MIGRATE` needs. **Deferred (2026-08-25)** per the user's own "we're going to have to