docs: frank-primary runs FreeToken (PR587 ROCm) — runbook, index, AGENTS, log

New runbook frank-primary-freetoken-pr587 (build laws, shape, measured
cutover numbers, TP=2 non-boot signature, rollback); LLM/index.md intro +
mermaid + active/retired entries; AGENTS.md fleet row + section + bench
and model-config mentions + runbook index; log.md entry. Seam pins the
measured KV pool 299968 with the multiple-of-64 guard.

NOTE: intended PGP signature skipped — key 0x76EA601A11B178D7 has no
secret key in this machine's keyring (fresh ~/.gnupg); the PGP identity
lives elsewhere.

tests/frank_primary_unit.bats

@@ -99,13 +99,17 @@ DOCKERFILE="$REPO_ROOT/scripts/docker/frank-primary-freetoken.Dockerfile"
 @test "the FreeToken engine shape is in the env file" {
     # KV pool: TOKENS, not bytes. 24 KiB/token (12 full-attention layers x
     # 2 KV heads x 256 head_dim x K+V x f16) — the 36 GDN layers use fixed
-    # per-seq state, not per-token KV. 300,000 tokens = 7.2 GiB and clears
+    # per-seq state, not per-token KV. 299,968 tokens = 7.0 GiB and clears
     # the harness worst case (140,492-token prompt + 36,864 max_tokens =
     # 177,356); one NATIVE 262,144-token conversation + full budget also
-    # fits. The arithmetic guard keeps future resizes honest.
-    grep -q '^FRANK_PRIMARY_FT_NUM_TOKENS=300000$' "$ENV_FILE"
+    # fits. The value must stay a MULTIPLE OF 64: the auto attention
+    # backend is qsa_sparse, which overrides page size to 64 (measured
+    # first boot — 300,000 was refused). The arithmetic guard keeps future
+    # resizes honest.
+    grep -q '^FRANK_PRIMARY_FT_NUM_TOKENS=299968$' "$ENV_FILE"
     nt=$(awk -F= '/^FRANK_PRIMARY_FT_NUM_TOKENS=/{print $2}' "$ENV_FILE")
     [ "$nt" -ge 177356 ]
+    [ $((nt % 64)) -eq 0 ]
     # First-boot conservative shape: 2 decode slots, 10% expert LRU
     # (~12 GiB of the ~120 GiB fp8 expert set), output ceiling = the
     # harness maxTokens, single R9700 at index 0.