[CLAUDE] Docs: S103 AI_INFRA 6-broadcast adap — crystallized-backfill (measure-then-fill) + Harness-20 fix-all-structure + memory-hygiene mapping

Adap 6 broadcast (hash MATCH cả 6 indep-recompute) qua 2-workflow Harness-9:
- Harness-20 fix-all-structure -> engine §L + dogfood model-note 4-tầng-sweep (incl PERSONA) + 220K→380K neo-supersede + 2-đầu-catch (review bắt 2 DẪN-XUẤT stale-trailer sweep-sót → fixed)
- F4 crystallized-backfill = NEW scripts/crystallized-backfill.ps1 (measure-then-fill DRY NO-API; default target=0=OFF owner-authority; calibrated-range; precondition; file-floor=UPPER-BOUND honest) + budget.json config + budget-table wire
- F1/F3 already-met (0 hardcoded-violation persona-clean; BƯỚC 0.6 bắt S103=Opus real-metadata)
- session-cmd budget-table (Up1) + run-loop 2-tầng (Up2 largely-already-met §L.b(c))
- H16 no-key Applied% = NEW scripts/applied-eval-nokey.ps1 (adversarial-gate fail-closed 3-ca + fault-inject FLIP; mechanism-built numbers-DEFERRED)
- memory-hygiene A/B already-met (archive-gate DRY-RUN + governance-detectors 4-layer 1:1); C n/a JS-sandbox no-CLI-launcher
- consolidation reference-ack (crystallization≠backfill)

engine §L+§M · 6 adap-report + 1 adap-request (#53-garble reasoning-lane mitigation) · send-email fc1e22ac selftest-MATCH · check-email 3 pulled+verified+STAGE-2.

2-workflow: WF1 implement wf_55ef6778-a0d (3 lane schema-CLEAN, selftest exit-0 ×2 + fault-inject teeth) + WF2 review wf_3c5c37cb-aa2 fable-clone (Lane-C PASS returned + Lane-B recover-from-memory PASS; 0-BLOCKING 2-CONCERN-addressed). #53 garble ×3 → em-main recover; investigator AS-10 garble-curation REVERTED (git-backstop).

Runtime S103 = Opus (Fable outage lại, per-session flip; H5-fallback revert-free). State GIỮ NGUYÊN: Mig 61 · 89 bảng · 440 test · gotcha 79 · menu 54. 0 prod-code.

Co-Authored-By: Claude Opus 4.8 (1M) <noreply@anthropic.com>

.claude/agent-memory/memory-budget.json

@@ -68,6 +68,22 @@
       "read_only_scope": "MFE WRITES exactly one file: .claude/agent-memory/.mfe-state.json (its own strikes-state for cross-run Goodhart compare). 'READ-ONLY' is scoped to token_governor / the budget caps (owner-authority) which it NEVER writes -- it is NOT the claim 'writes nothing'."
     }
   },
+  "crystallized_backfill": {
+    "_note": "Harness-20-era (S103, 2026-07-06) crystallized-backfill = 'measure-then-fill'. Read by scripts/crystallized-backfill.ps1 (deterministic NO-API DRY planner). Adapts AI_INFRA broadcast 2026-07-04 F4 + session-cmd-budget-display Upgrade-1. DISTINCT from CRYSTALLIZATION (engine §I distill = compress-TOWARD-source, keep time-invariant): backfill goes the OPPOSITE direction = FILL the empty hot-feed budget with already-crystallized cold-tier data so the cap is not wasted. PRECONDITION: expected_backfill opens ONLY when target>0 AND measured_headroom>0. Default target=0=OFF. NOT-for-sub (sub packs are lead-curated per-task; blind-pour dilutes relevance).",
+    "target": 0,
+    "_target_note": "OWNER-AUTHORITY static number (mirror token_governor.role_boundary_note + mark RC-...21-06...01-58-01). 0 = OFF/disabled = default. Owner raises >0 ONLY after judging a POSITIVE real-token headroom exists on the stable hot-load floor. The AI (em-main) NEVER sets/auto-tunes this; the script is READ-ONLY on it.",
+    "hotload_sources": [
+      "docs/STATUS.md",
+      "docs/HANDOFF.md",
+      ".claude/governance/ACTIVE-MARKS.md",
+      "docs/changelog/migration-todos.md",
+      "docs/gotchas.md",
+      ".claude/agents/README.md"
+    ],
+    "_hotload_note": "STABLE always-on FILE-FLOOR only (reproducible session-to-session). measured_headroom over this list = an UPPER BOUND on real headroom: the PEAK hot-feed ALSO loads session-VARIABLE lead-pasted task-context (task-relevant docs / code / gotcha-slices / engine-slices) that is NOT a fixed file-list, so the real peak approaches the ~368K lead read-set (token_governor.lead_note) and real headroom is SMALLER than this floor-measure shows. This is exactly broadcast F4's warning: byte/4 undercount + partial-source-list = 'illusion of room'. Hence target stays 0 until the owner judges stable room exists.",
+    "source_order": ["gist", "value-marked-archive", "curated-RAG"],
+    "_source_order_note": "value-gated, dedup vs hot-load, only TIME-INVARIANT crystallized content (NOT raw-episodic / volatile): (1) archive/<period>.gist.md 4-field distill; (2) archive lines carrying value_protect markers (gotcha# / anti-pattern / root-cause / lesson); (3) curated RAG search_memory. The PLANNER prints this order; it performs NO pull (DRY, like memory-archive-gate.ps1) - zero context change."
+  },
   "harness_floor": {
     "_note": "Harness-15 A1/A3 (S81, 2026-06-20): SAN-harness = fixed per-spawn cost (NOT tunable) = tool-schema + framing + own persona/role file + lead-pasted base-doc slice + task prompt. SEPARATE HOUSE (A3 anti-double-count): persona + lead-pasted-docs belong HERE (floor), NOT counted in token_governor.l1_always (which = own agent-memory + archive index + work-state block only). MEASURED-ESTIMATE not exact (H15 honest-note b): persona = directly measured bytes (.claude/agents/<name>.md 4.3KB-13.3KB => ~1.3K-4.0K tok via /3.3); tool-schema + framing = harness-injected (cannot byte-count locally), estimated comparable to AI_INFRA same-toolset-family (Read/Write/Edit/Bash/Grep/Glob/Skill/RAG).",
     "measured_token_estimate": 21000,