session 2026-07-08: V2 Fable scope charter (2 jurist passes) + MemPalace palace-memory decommission + BMF report + studium-cm-sandbox memory; REVIEWED-51 placed

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015jYS1tpR4P4NBxui4FSgfv

REVIEWED.md

@@ -426,3 +426,28 @@ DONE + DEPLOYED (a9e5196).
 - **C — acknowledged, in-loop:** `fidelity_equivalence@1` (REVIEWED-48) stands; the VIAF/ISNI/Wikidata reconciliation now targets the A1 Agent/Nomen registry specifically.
 
 **If AUTHORIZED:** Executor proceeds per the action-list (frontmatter sweep now-unblocked; build the Agent/Nomen registry, answering the `retrieve.py` hot-path question first; verify Olson then draft A2's preamble; draft A3 as a PROPOSAL; A4 after A2; B ordinary; C reconciliation-target noted). Every advanced item runs the RFC-supersession amendment process, each its own PENDING/REVIEWED. Tag related commits REVIEWED-49.
+
+## REVIEWED-50 — A2 Preamble Editor-Gate (advances REVIEWED-49 · A2)
+**Date:** 2026-07-07
+**Decision:** AUTHORIZED
+**Supersedes:** PENDING-50 · **Ruling type:** jurist editor-gate (format + doctrinal soundness), steward-ratified
+
+**Notes:** Editor-gate PASSED. The preamble says what REVIEWED-49 · A2 ruled and only that:
+- Berman carries the narrow claim (bias is real; own it rather than chase a false neutrality); Olson carries eccentric, *situated placement within a system* — not the self-standing-scheme claim; the bounded self-standing scheme is marked as the Chamber's own move, made *against* Olson's stated preference for permeability over new limits, attributed to no one. No A4 leakage.
+- **§2 (adopted):** a bridging clause opens the subsection, distinguishing *arrangement-honesty* (standpoint declared, not hidden) from §I's *fidelity* (checkable-against-source) sense of trust.
+- **§3 (resolved):** "a bounded one must" reads as conceptual entailment, not a tooling-enforced obligation — FIX-class holds.
+- **Placement CONFIRMED:** new subsection under §I. **Lane CONFIRMED:** v2.0.1 patch supersession (not errata — doctrinal content a successor must be able to trace to a dated marker; the legibility-to-successor value under A3).
+
+**If AUTHORIZED:** LANDED 2026-07-07 as spec **v2.0.1** — new §I subsection "On the partiality of the arrangement"; v2.0.0 archived immutable (`chamber-library-specification-v2.0.0.md`); chamber `CLAUDE.md` pointer updated; PENDING-50 closed; committed `34c39f1` (pushed github + skemantix). **Next:** A4's scope-citation redirect (now that A2 has a ratified home to point to), then the REVIEWED-49 tail. Tag related commits REVIEWED-50.
+
+## REVIEWED-51 — PENDING-51 — A4 + B(iii): the corpus-as-data framing term + the scope-warrant it does not carry (§I subsection)
+**Date:** 2026-07-07
+**Decision:** AUTHORIZED — jurist editor-gate PASSED; steward ratified. Advances REVIEWED-49 · A4 + B(iii). (Spec context: `chamber-library-specification.md`, patch supersession v2.0.1 → v2.0.2, OPERATIVE.)
+**Notes:** One coupled §I subsection — *'The corpus as data — and the warrant it does not carry'* — that introduces 'collections as data' as a **computational-tractability** framing term (B(iii)) and records, in the same breath, that this framing does **not** supply the corpus's bounded/personal scope warrant; that warrant is the §I partiality preamble (Berman/Olson, REVIEWED-50), not Santa Barbara Principle 4 (A4).
+
+- **Diagnosis (verify-against-substrate):** grep of the live spec for 'collections as data' returned ZERO hits — the phrase and the refuted Principle-4 scope-citation lived only in the research briefs, never the constitution. So **A4 was preventive/recording** (nothing to redirect in place) and **B(iii) was the term's first entry.** Collapsed to one subsection, the A4 boundary written *into* B(iii)'s introduction — honouring the RULING's "computational-only, no scope-justification leakage" flag structurally.
+- **Primary-source verification (steward-directed, before the gate):** the primary Santa Barbara Statement was fetched. **Principle 8 CONFIRMED verbatim** ('the technical integrity of the data as well as its provenance'; 'acknowledge absences and areas of uncertainty'). **Principle 4 CORRECTED** — it *does* concern designing for specific communities, but as **audience-fit of a data product**, not the **epistemic warrant for a bounded/personal curation**; the frontier brief's flat shorthand was imprecise, and the draft now states the precise distinction. The A4-shaped citation error was caught on our own draft by reading the source.
+- **Jurist editor-gate** (verbatim RULING: `chamber-library/docs/library-science-A4-Biii-RULING-2026-07-07.md`): PASSES 1/2/4/5 as drafted (Q2 "stronger than a minimal pass" — naming Principle 4 to refute it in the same breath is the safer construction; Q1 correction "the right level of precision"). **Q3 ruled confirm-don't-default-soften**; the executor confirmed by §IX substrate read → **Case A**: §IX's own `Still open` text settles the coverage-ledger, evidence-tier column, and freshness/honest-degradation, leaving ONLY silence-as-finding open (explicitly separated) → citation stands as drafted, **both terms, NO edit** (the one-word fallback was not triggered).
+- **Lane CONFIRMED:** v2.0.1 → v2.0.2 patch supersession (rationale text, no gate change, no consumer re-verification). Placement §I, immediately after the A2 partiality preamble.
+
+**If AUTHORIZED:** **LANDED** — committed `7d1d8d3` (local; Gitea push is the steward's call). `chamber-library-specification.md` bumped v2.0.1 → v2.0.2 (subsection added + v2.0.2 header supersession note); v2.0.1 archived immutable (`chamber-library-specification-v2.0.1.md`, byte-identical, `cmp`-verified); chamber `CLAUDE.md` version pointer updated. Bounded-diff verified (canonical change = exactly the header note + the new subsection; nothing else). Governance record: PENDING-51. **Closes A4 + B(iii)** of the REVIEWED-49 action-list. Remaining tail: **A1** registry (answer the `retrieve.py` hot-path question first), **A3** self-audit PROPOSAL, **B(i)/(ii)** (licence field · PREMIS vocab), **C** VIAF/ISNI/Wikidata reconciliation (→ targets the A1 registry). Tag any follow-on with REVIEWED-51.

claude/memory/MEMORY-reference.md

@@ -20,6 +20,9 @@ Split out of [MEMORY.md](MEMORY.md) on 2026-07-06 to keep the wake-loaded index
 
 # Archived sessions + stable reference layer (relocated verbatim from MEMORY.md, 2026-07-06)
 
+## Archived (2026-07-08 — Fable-window → Studium V2 charter + MemPalace decommissioned + BMF report; demoted on promote)
+- [Session 2026-07-07 evening — chamber A4+B(iii) & A1 landed; wake/wrap rewired off MemPalace](session-2026-07-07-evening-chamber-a4-a1-skill-rewire.md) — Post-/clear evening session; thread = **chamber-library**, held throughout. **Shipped 2 REVIEWED-49 items end-to-end, loop intact at every gate:** (1) **A4+B(iii)** → spec **v2.0.2** (the 'corpus as data — and the warrant it does not carry' §I subsection; diagnosis: the term had ZERO hits in the live spec → A4 preventive, collapsed with B(iii); **primary-source 'verify now' caught a Principle-4 imprecision on our own draft**; jurist editor-gate PASSED, Q3 confirmed by §IX read; steward-ratified; chamber `7d1d8d3`). (2) **A1 Agent/Nomen registry v1** (`_curation/agent-registry.yaml`; **134 agents / 143 strings / 222 of 1285 files — author coverage sparse ~18%**; steward's "how do serious libraries do this?" → adopted **library authority control**, IFLA-LRM/RDA 'attributed name'; generative `scripts/build_agent_registry.py`; Buddha=attributed; 5 compounds + 4 frontmatter defects flagged; chamber `703efcf`). Both **pushed to Gitea**. Then **rewired `/wake-up` + `/wrap-up` off palace-memory MemPalace → the files layer** (wind-down STAGED PLAN item 3; drift-patterns now grep `knowledge-graph.jsonl`; typography palace KEPT). **PULLING THREAD: the chamber REVIEWED-49 tail continues** (A3 PROPOSAL / B(i),(ii) / C reconciliation + A1 follow-ons) — beside it, the parked **MemPalace decommission + owed BMF lessons report** (the slot the steward is emotionally invested in — distil the loss into CapableMind-L1 design intent). REVIEWED-51 draft on Desktop awaiting placement.
+
 ## Archived (2026-07-07 evening — chamber A4+B(iii) & A1 landed + wake/wrap rewired off MemPalace; demoted on promote)
 - [Session 2026-07-07 — MemPalace wind-down decided (evidenced); KG secured; report owed](session-2026-07-07-mempalace-winddown-decided.md) — Wake opened on the REVIEWED-49 chamber folding but **PIVOTED entirely to MemPalace**. Diagnosed the "upgrade" = a custom **bge-m3 fork ~300 commits behind**, not a standard install; the morning "self-healed" drift = **108 chronic quarantines** (data intact: 88,159 rows). from-sqlite re-embed measured **~21 h**; dropping bge-m3 → a **quasi-standard install**. Steward **at limit** with MemPalace. **Independent 46-transcript audit: search+KG NOT load-bearing** (confirmatory/redundant; value migrated to files; graph features 0× used; write:read 7:1; conf 0.93/0.82). **DECISION (steward, evidenced): wind down `palace-memory` MemPalace; EXECUTE FRESH.** **KG exported+secured** → `knowledge-graph.jsonl` (329 triples) + `-entities.jsonl` (437). **BMF lessons report OWED** (reconnaissance framing; outline in `project-mempalace-winddown.md`). **PULLING THREAD next session: CHAMBER LIBRARY** (steward's choice) — resume the REVIEWED-49 folding (**A4+B(iii)** → A1 registry). Decommission + report = a separate 'tomorrow'; typography palace KEPT untouched.
 

claude/memory/MEMORY.md

@@ -43,6 +43,7 @@ permalink: claude-memory/memory
 - Chamber-typography — *tracker not yet established*; substantive moves live in per-session memories (2026-05-11 onward) + `project-chamber-cruft-restoration.md` + `project-chamber-typography-mining-plan-2026-05-15.md`.
 - [ARC chamber v1-legacy cluster](project-arc-chamber-v1-legacy-cluster.md) — ARC's `content/chamber/**` is intentional v1-Chamber legacy (not drift); deferred to-do = gather into a presentable cluster as a record of development. Out of §5 clause-1 audit scope. Confirmed 2026-05-29.
 - [Studium engine telos — the chamber of voices](project-studium-engine-telos-chamber-of-voices.md) — **the ultimate goal, hold above the build plan**: David's childhood imaginary chamber of beloved hero-voices he took counsel from → enter into discourse with his library + have the voices converse amongst themselves, this time *accountably* (the engine's v1 was literally "the Chamber" — eloquent/unaccountable simulacra; this is the same ambition built grounded/citable/checkable). Why verbatim fidelity is load-bearing.
+- [Studium = CM's unfettered sandbox](project-studium-cm-sandbox-and-transfer.md) — Studium/chamber are personal projects Seb now sees as fundamental to CM; experiment freely on the library/engine without risking CM's runtime, breakthroughs transfer back (V2 verifier = live example). Steward-framed 2026-07-08.
 - [Making sequence source set](project-making-sequence-source-set.md) — **COMPLETE against the ReadingList as of 2026-06-18** (reconciliation-verified); census of what's sourced where, sourced≠ingested (only Pos I in manifest), the **Handke-German decision** (original is source), the Levi drop-cap gate. Read before any Making/studium source work.
 - [Source library — link + dedupe](project-source-library-link-and-dedupe.md) — steward's master ebook library = `~/Documents/___The Library [ePub_AWZ3]/` (2190 ebooks, 180k files, messy nested; pre-organized subfolders `_2026 chamber source cleanup/`, `tmp calibre library for the chamber/`). GOAL: link chamber↔sources (provenance index) + dedupe. Seed = the 2026-06-29 fuzzy source-matcher (→ `_curation/reconvert-queue-2026-06-29.md`, 24/34 found). Real link needs EPUB/PDF internal metadata + edition-identity check, not just filenames.
 - [Character-as-image hazard](feedback-character-as-image-hazard.md) — some EPUBs render transliteration diacritics (ḥ/ʾ/ʿ) and non-Latin script (Hebrew words) as INLINE IMAGES; the standard image-drop SILENTLY MUTILATES them (PASS-BUT-FALSELY — gate passes, words corrupted). Tripwire = audit_cruft md_image count; INSPECT a sample before dropping; character-bearing images = CONTENT (map→Unicode, never drop). Found graduating Alter's Psalms (356 glyph-images) 2026-06-29; Alter parked pending a philological glyph-map (needs steward's Hebrew).
@@ -53,7 +54,7 @@ permalink: claude-memory/memory
 - [Be (laundromat)](project-be-laundromat.md) — canonical workstream tracker established 2026-06-08 (Seb-relay of locked decisions). Be = Skemantix startup (Seb+David) funding CapableMind's funding-ladder; **bridge, not venture**. Decisions LOCKED: entity/exit (CapableMind decoupled, grant-funded), pricing (Living $12.99/mo · Archive $69.99/yr · Memorial $49.99/yr · Renovate ~$199 · $8.99 floor), CF Self-Serve Agency + versioned-template-package infra. **a11y gate MERGED (Pat 100/100/100).** Pre-revenue: the WTP gate = renovate Pat → charge her. **Discipline: stop adding spec until the gate clears → nothing for executor on be until then.** Repo @ `f43a0fd`.
 
 ## Active Session
-- [Session 2026-07-07 evening — chamber A4+B(iii) & A1 landed; wake/wrap rewired off MemPalace](session-2026-07-07-evening-chamber-a4-a1-skill-rewire.md) — Post-/clear evening session; thread = **chamber-library**, held throughout. **Shipped 2 REVIEWED-49 items end-to-end, loop intact at every gate:** (1) **A4+B(iii)** → spec **v2.0.2** (the 'corpus as data — and the warrant it does not carry' §I subsection; diagnosis: the term had ZERO hits in the live spec → A4 preventive, collapsed with B(iii); **primary-source 'verify now' caught a Principle-4 imprecision on our own draft**; jurist editor-gate PASSED, Q3 confirmed by §IX read; steward-ratified; chamber `7d1d8d3`). (2) **A1 Agent/Nomen registry v1** (`_curation/agent-registry.yaml`; **134 agents / 143 strings / 222 of 1285 files — author coverage sparse ~18%**; steward's "how do serious libraries do this?" → adopted **library authority control**, IFLA-LRM/RDA 'attributed name'; generative `scripts/build_agent_registry.py`; Buddha=attributed; 5 compounds + 4 frontmatter defects flagged; chamber `703efcf`). Both **pushed to Gitea**. Then **rewired `/wake-up` + `/wrap-up` off palace-memory MemPalace → the files layer** (wind-down STAGED PLAN item 3; drift-patterns now grep `knowledge-graph.jsonl`; typography palace KEPT). **PULLING THREAD: the chamber REVIEWED-49 tail continues** (A3 PROPOSAL / B(i),(ii) / C reconciliation + A1 follow-ons) — beside it, the parked **MemPalace decommission + owed BMF lessons report** (the slot the steward is emotionally invested in — distil the loss into CapableMind-L1 design intent). REVIEWED-51 draft on Desktop awaiting placement.
+- [Session 2026-07-08 — Fable-window → Studium V2 charter; MemPalace decommissioned; BMF report](session-2026-07-08-fable-window-v2-charter-mempalace-decommission.md) — Opus session; the wake's chamber thread was **superseded** by a 3-day free-Fable-5 window (through **2026-07-12**). **Allocated it to Studium V2** (the multilingual/archaic entailment-validation harness) over ARC C1 — the steward's **microcosm insight**: Studium is CM's *unfettered sandbox* (code-separable, but breakthroughs transfer CM-ward → V2 pays twice; C1 = low-regret Opus/credits fallback). **Wrote + TWICE-jurist-gated the V2 Fable scope charter** (`studium-engine/docs/v2-validation-harness-scope-charter-2026-07-08.md`, `98d322c`; the /model-handoff discipline; §1 fence + §4 full open surface; the **œ/æ resolution** — exclusion re-derived from the ratified *closed* fold-list, not the executor flag — was the standout). **Fully decommissioned palace-memory MemPalace** (hooks + MCP server removed; all dead instances + the typography *backup* cold-archived to external `/Volumes/Temp storage 2020/…`; the **live typography palace KEPT**). **Wrote the BMF lessons report** (`CM-AI/…/l1-reliability/mempalace-decommission-report-for-bmf-2026-07-08.md`, `a5c7120`; CM index reconciled 35→38). New memory: [[project-studium-cm-sandbox-and-transfer]]. **⚠ PULLING THREAD = Studium V2; the NEXT session is BARE Fable (no /wake-up, no /wrap-up)** — the charter is its context, the design doc its only record; the thread points **past** it to the Opus resumption that jurist-reviews Fable's output (same seat), folds the jurist-gate-flags, and continues V2 build → V3.
 
 ## Historical reference → MEMORY-reference.md
 Older archived-session pointers and the stable reference layer (steward profile · project-state detail · L1/L2/Chamber inventories · legacy pending-work · reference-file list) live in [MEMORY-reference.md](MEMORY-reference.md) — consult on demand; not loaded at wake. Recent cross-session trajectory comes from the Active Session entry above + the recent `session-*.md` files (wake §2.b.1; the MemPalace `handoffs` glance was retired 2026-07-07 with the wind-down).

claude/memory/knowledge-graph.jsonl

@@ -331,3 +331,6 @@
 {"subject": "chamber-library", "predicate": "has", "object": "Agent/Nomen registry v1 (_curation/agent-registry.yaml; 134 agents; A1/REVIEWED-49)", "valid_from": "2026-07-07", "valid_to": null, "confidence": 1.0, "source_file": "session-2026-07-07-evening-chamber-a4-a1-skill-rewire.md", "extracted_at": "2026-07-07 22:28:46"}
 {"subject": "palace-memory-mempalace", "predicate": "status", "object": "winding down (steward 2026-07-07); wake+wrap skills rewired to files layer (STAGED PLAN item 3 done)", "valid_from": "2026-07-07", "valid_to": null, "confidence": 1.0, "source_file": "project-mempalace-winddown.md", "extracted_at": "2026-07-07 22:28:46"}
 {"subject": "claude-code", "predicate": "drift-pattern", "object": "invent-vocabulary-when-a-discipline-exists (steward 'how do serious libraries do this?' -> adopt authority control, not a homegrown scheme)", "valid_from": "2026-07-07", "valid_to": null, "confidence": 1.0, "source_file": "session-2026-07-07-evening-chamber-a4-a1-skill-rewire.md", "extracted_at": "2026-07-07 22:28:46"}
+{"subject":"claude-code","predicate":"drift-pattern","object":"verification-check-passed-vacuously (cross-vol move byte-sum computed 0=0; the real gate was file-count + rsync-rc + du — an instrument can pass without testing)","valid_from":"2026-07-08","valid_to":null,"confidence":1.0,"source_file":"session-2026-07-08-fable-window-v2-charter-mempalace-decommission.md","extracted_at":"2026-07-08"}
+{"subject":"mempalace-palace-memory","predicate":"status","object":"DECOMMISSIONED 2026-07-08 (hooks + MCP server removed; data cold-archived to external; typography palace KEPT live)","valid_from":"2026-07-08","valid_to":null,"confidence":1.0,"source_file":"project-mempalace-winddown.md","extracted_at":"2026-07-08"}
+{"subject":"studium-engine","predicate":"relationship-to-capablemind","object":"unfettered experimentation sandbox; code-separable (zero L1/BMF imports) but breakthroughs transfer CM-ward (V2 = live example)","valid_from":"2026-07-08","valid_to":null,"confidence":1.0,"source_file":"project-studium-cm-sandbox-and-transfer.md","extracted_at":"2026-07-08"}

claude/memory/project-mempalace-winddown.md

@@ -25,11 +25,20 @@ Independent audit of 46 sessions: MemPalace's **search + KG are NOT load-bearing
 3. **Rewire skills to files** — **✓ DONE 2026-07-07** (steward-directed). `/wake-up` §2.b → files-layer reconstruction (recent `session-*.md` trajectory + `grep drift-pattern knowledge-graph.jsonl`), §2.e upgrade-check retired, §5 constraints inverted (files primary); `/wrap-up` §4 → the session file IS the record, §5 → append JSONL triples to `knowledge-graph.jsonl` (soft-invalidate via `valid_to`), §8 template updated. No live `mempalace_*` calls remain; typography palace KEPT; both skills carry provenance notes; recorded in the skill-harvest register.
 4. **Cold-archive** the backed-up palace; delete the live churning one after verification.
 
+## EXECUTED — 2026-07-08 (items 1, 2, 4 — all reversible; hard deletion deferred)
+- **Item 1 (hooks) DONE:** removed the Stop (`mempal_save_hook.sh`) + PreCompact (`mempal_precompact_hook.sh`) hooks and the 3 `MEMPALACE_*` env vars from `~/.claude/settings.json`. Re-validated. Backup: scratchpad `mempalace-decommission-backups-20260708/settings.json.bak` (settings.json is NOT dotfiles-tracked — backup is the rollback).
+- **Item 2 (MCP server) DONE:** removed the `mempalace` block from `~/.claude.json` `mcpServers` (it pointed at palace-memory). Now `mcpServers = {plane}`; 0 `mempalace-mcp` refs; JSON re-validated. Backup: `.../.claude.json.bak`. Typography palace is CLI-accessed (not via this server) → unaffected; the mempalace install + venv at `~/_Dev/mempalace` KEPT for the typography CLI.
+- **Item 4 (cold-archive) DONE (reversible move):** `palace-memory` (838M) + `palace-memory.pre-rebuild-bak-20260707-1230` (824M) → `~/.mempalace/_decommissioned-palace-memory-20260708/` (1.7G) + README. **Hard deletion DEFERRED** to explicit steward confirm (reclaims ~1.66 GB). Root-level shared config + other palaces (`palace`, `palace-vault`, typography ×2) left untouched.
+- **This session:** mempalace MCP tools remain loaded (server process still live this session); from the next post-clear session there are no mempalace hooks and no mempalace server. An old Stop hook, if still cached this session, errors harmlessly.
+- **Item 4 EXTENDED → external cold storage (2026-07-08, steward-directed):** all dead instances relocated OFF the internal disk to `/Volumes/Temp storage 2020/_dev/mempalace decommission/` (3.1G; rsync rc=0 + exact file-count match per dir, sources removed after verify): `_decommissioned-palace-memory-20260708` (palace-memory + its backup), `palace-vault`, `palace`, `archive-chamber-library-2026-04-30`. Internal `~/.mempalace` now holds ONLY the typography palace + its backup + root-level shared config. → the earlier "1.66 GB hard delete" is moot (moved, not copied; the external copy IS the cold archive).
+- **BMF lessons report → DONE 2026-07-08:** `~/_Dev/CapableMind-AI/docs/thinking/David/l1-reliability/mempalace-decommission-report-for-bmf-2026-07-08.md` — closing chapter of the substrate-failure inquiry; mapped to Logchain-as-truth (#1) / honest-degradation (#4) / PENDING-16 (observation-recall) / PENDING-17 (epistemic integrity).
+- **Still owed (minor):** optional tidy of the inert `mcp__mempalace__*` allow-entries in `settings.local.json` (harmless without the server); the CM `thinking/README.md` index row for the report + its eventual commit (steward — shared repo w/ Seb).
+
 ## PRESERVE — do NOT touch
 - **Typography palace** (`~/.mempalace/palace-chamber-typography`, queried via CLI for ARC type work, [[reference-typography-palace-cli]]) — SEPARATE instance, genuinely used, KEEP as-is. The audit verdict is palace-memory ONLY. Do **not** do a global embedder flip (would break it, 1024-dim/bge-m3).
 - The exported drift-patterns — wire wake to read them.
 
-## BMF lessons report — OWED (steward-requested 2026-07-07), part of decommission (tomorrow)
+## BMF lessons report — DONE 2026-07-08 (`mempalace-decommission-report-for-bmf-2026-07-08.md`)
 **Framing (steward's own):** MemPalace was a long experiment — stopgap AND *reconnaissance*: a substrate close to our own vision, run to see **from the inside** how others approach the AI-memory problem. The report harvests everything of value **FOR BMF** (CapableMind L1, the real replacement), not a mere obituary.
 **Home:** `~/_Dev/CapableMind-AI/docs/thinking/David/l1-reliability/mempalace-decommission-report-for-bmf-2026-07-07.md` — build ON `substrate-failure-lessons-2026-06-07.md`, do not duplicate.
 **Sources to gather (fast):** substrate-failure-lessons-2026-06-07.md + sibling l1-reliability post-mortems (recall-quality-baseline, ingestion-forensic, deep-audit-plan); MemPalace's own `~/_Dev/mempalace/CLAUDE.md` design principles; THIS session's forensics (`session-ledger-2026-07-07.md` + the audit verdict); PENDING-16 (observation-recall coupling) + PENDING-17 (epistemic integrity) = BMF design intent to map to; the MemPalace feedback memories.

claude/memory/project-studium-cm-sandbox-and-transfer.md

@@ -0,0 +1,16 @@
+---
+name: project-studium-cm-sandbox-and-transfer
+description: "Studium engine + chamber-library are the steward's personal projects AND the unfettered experimentation sandbox for CapableMind — separable from CM's runtime, but breakthroughs transfer CM-ward; Seb now sees them as fundamental to the shared vision."
+metadata: 
+  node_type: memory
+  type: project
+  originSessionId: c08295e8-eaf6-4486-bb1a-9311c7cbe644
+---
+
+The Studium engine + chamber-library are David's **personal** projects — but Seb now recognizes them as **fundamental to the shared CapableMind vision**. The strategic value of keeping them separate: David can **experiment unfettered** on the library/engine without risk of breaking CM's production L1/BMF runtime — while breakthroughs still **transfer CM-ward**. The engine is a *microcosm of CM*: the same governance thesis (executor/jurist/steward loop, L2-over-L1, honest-degradation / earned-confidence) instantiated over a bounded, provenanced library instead of the L1 runtime. Code-separable (zero L1/BMF/`mcp__capablemind` imports; verified 2026-07-08); coupled only at the governance-thesis level.
+
+Concrete transfer surface — the V2 verifier work is the live example: per-language honest-degradation semantics ↔ CM's honest-degradation contract (REVIEWED-22); abstain-when-ungrounded ↔ earned-confidence ("the system does not grant epistemic authority to its own outputs without external grounding", REVIEWED-19); per-language trust/revise/abstain thresholds against a real failure taxonomy ↔ the confidence-floor/ceiling invariants (I-CF / I-CC), which L1 has only specified abstractly. Studium is where CM's epistemic-integrity invariants get pressure-tested on a real corpus before L1 hardens them.
+
+**Why:** this is the rationale for concentrating effort on the library/engine — a safe, unfettered ground for the epistemic risks CM's live runtime cannot take, with a transfer path back. Named by the steward 2026-07-08 (the Fable-window allocation conversation).
+
+**How to apply:** treat Studium/chamber as the experimentation sandbox (design freely, take research-grade risks there); when designing there, actively surface what transfers to CM — cf. the V2 charter's Deliverable 9, the CM-transfer note (`studium-engine/docs/v2-validation-harness-scope-charter-2026-07-08.md`). Keep them distinct projects — don't conflate Studium with CM, kin to [[mempalace-is-unaffiliated-stopgap]]'s build-vs-use discipline — but do harvest the transfer. Companion to [[project-studium-engine-telos-chamber-of-voices]] and [[reference-studium-engine-architectural-charter]].

claude/memory/session-2026-07-08-fable-window-v2-charter-mempalace-decommission.md

@@ -0,0 +1,56 @@
+---
+name: session-2026-07-08-fable-window-studium-v2-charter-mempalace-decommissioned-bmf-report
+description: "Opus session that allocated the 3-day free-Fable-5 window (through 2026-07-12) to Studium V2 (the verifier-validation harness) over ARC C1, wrote + TWICE-jurist-gated the V2 Fable scope charter, fully decommissioned the palace-memory MemPalace (hooks/MCP/data → cold storage), and wrote the BMF lessons report. Pulling thread: the Studium V2 verifier-validation work — the NEXT session is a BARE Fable-on-V2 run (no wake/wrap; the charter is its context, the design doc its only record), so the thread points PAST it to the Opus resumption that reviews Fable's output (jurist same-seat), folds the jurist-gate-flags, and continues V2 build → V3."
+metadata: 
+  node_type: memory
+  type: project
+  originSessionId: c08295e8-eaf6-4486-bb1a-9311c7cbe644
+---
+
+# Session 2026-07-08 — Fable-window allocation → V2 charter; MemPalace decommission; BMF report
+
+Opened by /wake-up → Symmetria init (thread confirmed = chamber REVIEWED-49 tail; the smaller literal question answered on entry — the steward had placed REVIEWED-51). The session then **superseded** that thread: the steward surfaced a **3-day free-Fable-5 promo window** (extended through Sun 2026-07-12) and asked where — *apart from CapableMind* — to spend Fable's design/architecture power before it goes credit-only. The whole session became allocation + prep + housekeeping; chamber work did not advance.
+
+## PAST — what we did + why
+
+**1. Fable-window allocation → Studium V2 (over ARC C1).** Two Explore surveys grounded it. ARC's live design-phase candidate = **C1 print/PDF typographic posture doctrine** (arc-typography.sty v0.2→v0.3; the Vignette turned out design-*complete*, Opus-build). Studium's = **V2, the multilingual/archaic entailment-validation harness** (rebuild-plan §2.4). The steward's **microcosm insight** decided it: Studium is CM's *unfettered sandbox* — separable at the code level (zero L1/BMF imports, verified) but the same governance thesis, so breakthroughs transfer CM-ward. That makes V2 pay twice (engine + CM epistemic-integrity de-risking), and C1 the low-regret fallback (modest, narrow — Opus lands it fine, or a small credit spend). **Decision: V2 first, metered, bounded; C1 falls back to Opus/credits.**
+
+**2. The V2 Fable scope charter — written + TWO jurist editor-gate passes folded.** `studium-engine/docs/v2-validation-harness-scope-charter-2026-07-08.md` (committed `98d322c`). Built on the **/model-handoff** discipline (charter on cheap Opus → Fable spends premium tokens only on design; artifacts-not-chat; NO /wake-up on Fable — the charter IS the wake). §0 operating discipline + jurist-gate guardrail; §1 fence (settled — DO NOT re-derive); §4 the full open design surface in **hard priority order** (D1 claim-decomposition, D2 cross-lingual entailment, D4 verifier failure-taxonomy = DEEP; D5/6/7 core; D8 orthography; **D9 CM-transfer** graded grounded-or-abstain); §5 report-and-stop → one pre-registration-complete design doc. Two jurist passes closed on actual grounds: pass-1 (verdict-register guardrail §0.c, ß/ss-as-jurist-gate, D9-reflexive, budget-realism, provenance-watch); pass-2 (**œ/æ**, D1↔§0.c alignment, §5 stop-criteria, honesty-fix of an arranged-ellipsis quote). **Jurist signed off.** Carry-forwards (non-blocking): (a) a REVIEWED entry is owed (V0 method-gate never formally filed + a one-line œ/æ ratification — steward writes ~/REVIEWED.md); (b) **the jurist reviews the V2 design doc Fable produces, same seat** (recorded charter §5).
+
+**3. MemPalace `palace-memory` FULLY decommissioned** (steward-directed; the wind-down decided 2026-07-07). Items 1/2/4 executed, all reversible: hooks removed from `~/.claude/settings.json`; the `mempalace` MCP server removed from `~/.claude.json` (now `{plane}` only; JSON re-validated; config backups in scratchpad `mempalace-decommission-backups-20260708/`); all dead instances (palace-memory + backup, palace-vault, palace, archive-chamber-library, **and** the typography-palace *backup*) cold-archived to `/Volumes/Temp storage 2020/_dev/mempalace decommission/` (~3.6G, verify-before-delete: rsync rc=0 + exact file-count match, sources removed). **The live typography palace + the mempalace install/venv are KEPT** (typography CLI depends on them). Tracker: [[project-mempalace-winddown]].
+
+**4. BMF lessons report written** — `CM-AI/docs/thinking/David/l1-reliability/mempalace-decommission-report-for-bmf-2026-07-08.md` (committed `a5c7120`). Reconnaissance-not-obituary: MemPalace failed at *our* thesis (verbatim/local-first/100% recall) with *our* commitments, so its 4-month decay is empirical proof of the constitution's Logchain-as-truth (#1) + honest-degradation (#4) + PENDING-16/17; the files-authoritative convergence (its KG survived only by *becoming a file in git*). Closing chapter of the substrate-failure inquiry. **Also reconciled the CM thinking/README.md index** (David/l1-reliability 35→38: this report + 2 pre-existing un-indexed files).
+
+**5. New durable memory:** [[project-studium-cm-sandbox-and-transfer]] — the microcosm/sandbox framing, so it holds across sessions.
+
+**6. chamber-library README trued up to v2.0.2** — added the A1 Agent/Nomen registry (layer + `build_agent_registry.py`) + the REVIEWED-49 tail to known-open-work; kept the tier-counts honestly dated (no fake census). Committed `4c68cb6` (github + gitea).
+
+## Decisions explicitly NOT made (negative space)
+- **Chamber REVIEWED-49 tail did NOT advance** (A3 self-audit PROPOSAL, B(i)/(ii), C VIAF/ISNI reconciliation, A1 follow-ons) — the Fable-window detour took the session. Still open; now sits *behind* the V2/Fable work.
+- **ARC C1 not started** — deliberate: it's the modest fallback; the free-Fable window goes to the higher-leverage V2. C1 gets Fable only if budget remains, else Opus or credits.
+- **The V2 charter is PREP, not the design** — Fable does the design next.
+- **The owed REVIEWED entry not filed** — steward writes ~/REVIEWED.md; non-blocking.
+- **settings.local.json inert `mcp__mempalace__*` allowlist entries left** — harmless without the server; optional tidy.
+
+## PRESENT — mood / disposition
+Allocation discipline + verify-against-substrate throughout; the loop held at every boundary. The two best moments were both substrate-over-authority: (a) the **microcosm insight** (steward's) reframing V2 from engine-work to CM-transfer; (b) the **œ/æ resolution** — re-deriving the exclusion from the *ratified closed fold-list* ("…and nothing else") rather than accepting the jurist's "probably gate it" guess OR leaning on the executor cross-cutting flag as if it were ratified (the jurist called it the strongest reasoning in the exchange, and it's the exact item-3-launder the chamber exists to prevent). Honest small catch: my own cross-volume-move verify script's byte-sum computed `0=0` **vacuously** — the real proof was file-count + rsync-rc + du; caught and reported rather than trusted (kin to green-gate-reads-as-thoroughness).
+
+## FUTURE — what is pulling
+
+**PULLING THREAD: the Studium V2 verifier-validation harness** — the multilingual/archaic entailment-validation, now Fable's to design against the twice-gated charter.
+
+**⚠ NEXT SESSION IS BARE FABLE — no /wake-up, no /wrap-up** (steward-directed 2026-07-08). It opens directly on `studium-engine/docs/v2-validation-harness-scope-charter-2026-07-08.md`, executes §4, produces the design doc, and stops. **Fable leaves NO session memory** — the design doc is its only record.
+
+**ACTIONABLE RESUMPTION POINT (for the Opus session AFTER Fable — re-judge against what changed):** read `studium-engine/docs/v2-validation-harness-design-<date>.md` (Fable's output = the record of what it did); **jurist-review it, same seat**; fold its **jurist-gate-flags** (verdict-register / decomposition-failure fork; any ß/ss-or-quote/em-dash proposal; adversarial-class/proportional-distribution touches) + the **design-decisions** note; then hand to Opus for the V2 *build* + benchmark run, then V3. If the window ran hot and C1 wants doing, that's the parallel ARC track (Opus/credits).
+
+**Other open horizons (ranked):**
+- Chamber REVIEWED-49 tail (A3 PROPOSAL jurist-gated · B(i) licence field · B(ii) PREMIS · C VIAF/ISNI → the A1 registry · A1 follow-ons: 5 compounds, 4 frontmatter defects, test_tools integration).
+- ARC C1 (print/PDF posture doctrine) — Fable-if-budget, else Opus/credits.
+- Owed REVIEWED entry (V0 gate + œ/æ one-liner); settings.local.json tidy.
+- Older parked: PENDING-43 ESCALATE Loom/Mill; PENDING-27.
+
+**PAUSE STATEMENT:** Stepping away with a genuinely clean board — 3 project repos committed+pushed, MemPalace retired and cold-archived, the V2 charter twice-gated and Fable-ready. What I want to find still pulling: the V2 verifier work as the live thread, and the discipline that the *bare* Fable session's output gets the same jurist eye + honest grading the charter itself received — the charter is only as good as the review of what it produces.
+
+**LITERAL QUESTION for the post-Fable Opus session:** Did the Fable session produce the V2 design doc as scoped — and does its completeness-critic pass or jurist-gate-flags surface anything that reshapes the V2 build (a charter gap we didn't anticipate, or a (language, tier) cell that gated-to-abstain and thereby changes the D9 CM-transfer grading)?
+
+**State at wrap:** chamber-library `4c68cb6` (github+gitea), studium-engine `98d322c` (github), CapableMind-AI `a5c7120` (github) — all pushed. dotfiles/memory committed+pushed at this wrap. MemPalace palace-memory decommissioned + cold-archived; typography palace live. Fable free-window open through 2026-07-12.

claude/memory/session-ledger-2026-07-08.md

@@ -0,0 +1,42 @@
+---
+name: session-ledger-2026-07-08
+description: "Practice-of-return ledger maintained by /symmetria — returns, open horizons, recalibrations, authorization moves, sub-agent dialogues, bypasses."
+metadata: 
+  node_type: memory
+  type: feedback
+  originSessionId: c08295e8-eaf6-4486-bb1a-9311c7cbe644
+---
+
+# Session Ledger — 2026-07-08
+
+## Returns
+
+## Open horizons
+- 2026-07-08T16:45 — Woke into the **chamber REVIEWED-49 tail** (thread CONFIRMED). Smaller literal question answered on entry: REVIEWED-51 IS placed (REVIEWED.md:443–454, AUTHORIZED). Main fork still open for the steward: continue the chamber tail (lightest = B(i)/(ii) or the A1 compound follow-on) vs take the parked MemPalace-decommission + BMF-lessons-report slot.
+- Chamber tail inventory: A1 follow-on (5 compounds_pending → component Agents w/ roles) · B(i) licence field · B(ii) PREMIS name · A3 self-audit (PROPOSAL, jurist-gated) · C VIAF/ISNI/Wikidata reconciliation · test_tools.py integration · 4 frontmatter defects → sweep.
+
+## Fable-window allocation (3-day promo, through 2026-07-12; 50% weekly limit, draws down fast)
+- 2026-07-08T~17:00 — Steward asked: apart from CapableMind, where to spend Fable's design/thinking power before it goes credit-only. Two Explore surveys grounded it (ARC; Studium/chamber). Findings:
+  - **ARC C1 — Print/PDF posture doctrine** = the clean primary. arc-typography.sty is v0.2 (2026/05/14); §II.e never authored; print justify-vs-soft-rag / measure geometry / quote-norm / heading weight / glyph routing all UNDECIDED. Resolvable on paper, hard-to-reverse (governs all future ARC PDFs incl. Lune's translations), unblocked, unambiguously non-CapableMind. Vignette is design-COMPLETE (Opus build). Making essay = research/authorial (steward-owned).
+  - **Studium V2 — multilingual/archaic entailment-validation design** = highest engine leverage ("single most important artifact this phase"); genuine research-grade reasoning (does NLI entailment survive archaic multilingual humanities prose). Unblocked, D-1, NOT jurist-gated (V0 gate already cleared, REVIEWED-48). BUT engine is governance-thesis-adjacent to CapableMind (separable at code level; steward's scope call) AND wants experiments, not pure paper-design.
+  - **HOLD (discipline, not neglect):** Be = spec LOCKED until WTP gate clears (frozen since 06-07) — Fable window is NOT a reason to break it. Chamber-of-voices protocol schemas = deliberately DEFERRED until corpus-gold + one trustworthy voice; using the window to break that sequencing = executor-bias-to-compose. Named, not taken.
+  - Meta-move recommended: prep the scope charters on Opus FIRST (cheap orientation, /model-handoff discipline) so each Fable session spends premium tokens only on content reasoning.
+- RESOLVED (steward 2026-07-08): Studium IS separate (+ microcosm-of-CM framing → [[project-studium-cm-sandbox-and-transfer]]); **V2 first, metered, bounded** (design + pre-register the validation harness); C1 fallback to Opus/credits (modest). **V2 charter WRITTEN + jurist editor-gate FOLDED** (`studium-engine/docs/v2-validation-harness-scope-charter-2026-07-08.md`): 3 real fence-gaps closed — verdict-register guardrail §0.c; ß/ss + quote/em-dash marked jurist-gate not mechanical (§1.2/D8); D9 now grounded-or-abstain-reflexive (grade each transfer) — plus budget-realism + provenance-watch (§0) + §3 path confirmed present. Fable-ready. **Next: steward wraps → clears → Fable on the charter (NO /wake-up).**
+- **V2 charter JURIST-SIGNED-OFF 2026-07-08** — two editor-gate passes, all findings closed on actual grounds (the strongest being the **œ/æ resolution: excluded by entailment from the ratified closed fold-list, NOT by the executor flag** — no item-3 launder; the jurist called it the strongest reasoning in the exchange). **Carry-forwards (non-blocking):** (a) a REVIEWED entry is owed — the V0 method-gate was jurist-cleared but never formally filed, + a one-line œ/æ ratification would keep the trail clean (steward writes ~/REVIEWED.md; hygiene, not a gate); (b) **the jurist will review the V2 design doc Fable produces (same seat)** before it governs the build — recorded in charter §5.
+- **NEXT SESSION = BARE Fable research on V2 ONLY — no /wake-up, no /wrap-up** (steward-directed 2026-07-08). Continuity implication: **Fable will leave NO session memory** — its artifact/record IS the V2 design doc it produces (`studium-engine/docs/v2-validation-harness-design-<date>.md`, per charter §5). So the next *Opus* wake (the session AFTER Fable) resumes by reading that design doc as the evidence of what Fable did, reviewing it (jurist same-seat), folding jurist-gate-flags, and continuing to V3. This wrap's pulling-thread therefore points **past** the Fable session to that Opus resumption.
+
+## Confidence to recalibrate
+- chamber `703efcf → a095f7b` landed AFTER the documented wrap (docs: build_agent_registry.py to fleet). Verified via git log; synced to origin. Not authored this session — flag it as world-moved, don't assume it was mine.
+
+## Authorization moves
+- REVIEWED-51 placed by steward (dotfiles `M REVIEWED.md`, uncommitted — sweeps at wrap, NOT at wake).
+- **MemPalace decommission EXECUTED (steward-directed) 2026-07-08:** hooks removed (settings.json), MCP server removed (.claude.json), all dead instances moved to external `/Volumes/Temp storage 2020/…` (3.1G, verify-before-delete), typography palace kept. Config backups in scratchpad. Reversible throughout; no ESCALATE (MemPalace = unaffiliated stopgap). Tracker [[project-mempalace-winddown]] updated.
+- **BMF lessons report WRITTEN** → `CM-AI/docs/thinking/David/l1-reliability/mempalace-decommission-report-for-bmf-2026-07-08.md`. The reconnaissance dividend: MemPalace failed at OUR thesis with OUR commitments → empirical proof of Logchain-as-truth + honest-degradation (#4) + PENDING-16/17; the checklist of what "honest degradation" must build. NOT committed (shared repo, steward's call). CM thinking/README.md row owed.
+
+## Sub-agent dialogues
+
+## Bypasses
+
+## Drift flags to hold today
+- invent-vocabulary-when-a-discipline-exists (2026-07-07) — live risk: A3/C touch authority-control vocabulary. Reach for the established discipline (IFLA-LRM/RDA/VIAF), don't homebrew.
+- green-gate-array-reads-as-thoroughness (2026-07-06) — a row of PASS is not coverage.