dotfiles/REVIEWED.md

REVIEWED.md — Steward Authorization Record

Protocol: David and Claude.app write here after reviewing PENDING items. Claude Code reads this at session start.

---

REVIEWED-1 — Executor Agency Directive for CLAUDE.md

Date: 2026-03-21 Decision: AUTHORIZED Notes: Accepted as written with one addition: insert explicit one-sentence description of the specific failure mode ("accept self-assessment of criticality") immediately after the contamination problem reference. Steward to apply to ~/CLAUDE.md. This proposal is itself evidence the directive is already operative — the executor used the authorization taxonomy correctly on a change affecting its own behavior.

REVIEWED-2 — Silent Degradation Audit

Date: 2026-03-21
Decision: AUTHORIZED — commence after PR merges. Option 1 (full audit, concentrated). Produce findings document for steward review before any implementation.

REVIEWED-3 — Factory Schema Idempotency (checkpoint.ts, dedup.ts)

Date: 2026-03-21 Decision: ACKNOWLEDGED — already in PR scope. The executor's decision to extend the fix to the actual error source (checkpoint.ts) rather than the originally specified file (job-store.ts) is the correct application of "diagnose the class, not the instance." Note this in PR description.

REVIEWED-11 — Approve I15 (ICP-9 Pilot Registry Entry)

Date: 2026-03-23 Decision: AUTHORIZED Notes: I15 approved as pilot registry entry. l1_contamination_profile field ratified as mandatory for all non-contingent principles. Schema commitment is prospective — all future Cluster A/B entries must include this field. residual_risk field decision deferred pending evidence from pilot operation.your annotation]

REVIEWED-12 — Lodge Design Notes DN-GOV-01 through DN-GOV-04

Date: 2026-03-23 Decision: AUTHORIZED Notes: DN-GOV-03 requires separate declaration before elevation. DN-GOV-04 is tooling, iterate freely.

REVIEWED-13 — DN-GOV-05, DN-GOV-06, DN-GOV-07 (Threshold Inquiry Cluster)

Date: 2026-03-28 Decision: AUTHORIZED with jurist additions Notes:

DN-GOV-05 (Bounded Self-Repair Principle): Authorized as DESIGN COMMITMENT. Four-condition framework (reversibility, within parameters, independent verifiability, steward-legible framing) is sound. Suspension clause: enforcement mechanics not active until AF-7 reaches design commitment status. The principle is named and load-bearing in constitutional reasoning; it does not authorize autonomous self-repair actions until external review is constitutionally specified.

DN-GOV-06 (Temporal Authorization Shift): Authorized as DESIGN COMMITMENT. Mechanism constraints (never reaches amendment, never creates new authority, always conservative direction, no compounding) are correctly specified. Suspension clause: same as GOV-05 — not active until AF-7 reaches design commitment status. Additional constraint (jurist): temporal shift may not be invoked to justify actions that would not pass the GOV-05 four-condition test in a non-time-pressured context. Time pressure is a trigger, not a relaxation of conditions.

DN-GOV-07 (The Threshold Already Crossed): Authorized as FRAMING DOCUMENT. The "threshold already crossed" observation is entered into the constitutional record as an accurate description of current practice. It does not authorize further crossings. Two additions:

1. Working framing ("conditions of trustworthy service") is explicitly provisional — may be revised by L1 maturation evidence, Chamber Phase 3+ evidence, or steward philosophical judgment.
2. Counter-norm on inherited legitimacy: the fact that a threshold has been crossed does not create precedent for further crossings. Each additional autonomy claim requires fresh warrant, not inherited legitimacy from the prior crossing.

Cross-cluster consequence: AF-7 (External Review) is escalated from resolution brief to urgently load-bearing. AF-7 reaching design commitment status is the condition for the GOV-05/06 suspension clause to lift. AF-7 design note should be the next constitutional work item after I17.

REVIEWED-14 — DN-GOV-08: Constitutional Stabilization, Not Automation of Recognition

Date: 2026-03-28 Decision: AUTHORIZED — entered as ACTIVE constitutional record Notes: L2 constitutes boundary conditions for recognition; does not automate it. I3 is structural commitment, not capability limitation. Evaluation test added: any invariant touching recognition must be assessed against this commitment before hardening. Phase 3 placeholder (Talmudic chain of transmission, Traditionis) recorded — primary texts not yet in library, blocking condition explicit. Source: Chamber Phase 1, Essay I.

REVIEWED-15 — I17: Precedence of Present Expression

Date: 2026-03-28 Decision: AUTHORIZED Notes: Full registry entry following I15 pilot schema. Four-source convergence (Levinas, Fricker, Leopardi, Harrison). Anti-model-disambiguation clause incorporated per adversarial review. Jurist's design question resolved constitutionally: the system does not distinguish genuine divergence from momentary aberration — the interlocutor does. DN-GOV-08 evaluation test passed: I17 holds open the space for recognition, does not generate a recognition-outcome. Jurist's required amendment (RA-I17-01) incorporated before authorization: noting-frequency drift markers, noting-as-model-assertion constraint, noting threshold as human-declaration requirement. I3 category limit explicitly connected to the correct-model/wrong-expression asymmetry. Forward dependency to vector module governance amendment threaded. L1 contamination profile: critical, monotonic.

REVIEWED-16 — I16: The Asymmetry Obligation

Date: 2026-03-28 Decision: AUTHORIZED Notes: Five-source convergence (Harrison, Vico, Leopardi, Levinas, O'Neill) — strongest in the registry. Survived adversarial review without modification (priority 1). DN-GOV-08 evaluation passed: metrics function as monitoring surfaces, not obligation-generators (the score triggers review, not action). Two pre-authorization amendments incorporated: RA-I16-01 (obligation registration validity — closes FM-R4 at point of occurrence, not deferred to drift marker) and SD-I16-01 (capability_differential estimated from behavioral signals only — tracks engagement, not identity). Experienced-interlocutor prohibition absolute: familiarity deepens obligation, never relaxes it. Service-as-dominion detection marker included (anticipatory behavior correlated with asymmetry depth). L1 contamination profile: critical, monotonic — the contamination looks like good service. Cluster A now complete (I15, I16, I17). Post-authorization action: AF-7/C-R2 external audit mechanism advances immediately. The compound failure (I16 + I17 = self-confirming system) has monitoring but no enforcement until AF-7 exists.

REVIEWED-19 — Epistemic Integrity (PENDING-17)

Date: 2026-04-03 Decision: AUTHORIZED

Proposal to make classification confidence a load-bearing signal throughout the L1 pipeline. Currently computed and discarded — no module reads it, no store records it, no recall path weights by it. The system returns degraded guesses with the same authority as understood knowledge.

Authorized as [PROPOSAL] with the following commitments:

1. Constitutional position: Earned Confidence. "The system does not grant epistemic authority to its own outputs without external grounding." Assessed by jurist as constitutional position for L2 preamble. This is the normative claim from which the three invariants derive. The grounding chain terminates outside the system — ICP-19 and external review are constitutive, not decorative.

2. I-CF: Processing Confidence Floor (Cluster A). No module shall process an event whose classification confidence has not been earned against a declared floor. Sub-floor events are HELD for remediation via DeferrableError at base.ts:83, not discarded. Floor value to be calibrated from empirical data. The invariant is that the check EXISTS.

3. I-CC: Classification Confidence Ceiling (Cluster A). No classification confidence shall exceed the validated accuracy of the source that produced it. Unvalidated rules (accuracy null) cannot produce uncapped confidence. Enforcement by construction in classification.ts. Open schema question on enforcement vocabulary (CAP/BOUND) acknowledged — does not block implementation.

4. I-NF: Novelty Floor (Cluster A). Already authorized in REVIEWED-18. Confirmed.

5. Dual bloom filter. Understood vs degraded fingerprints tracked separately. Degraded content re-attended on re-encounter (retry opportunity, not penalized as duplicate). Both filters clear on repair/factory reset. Closes the permanent lockout kill chain.

6. Confidence provenance on stored records. Vector chunks, entity nodes, temporal nodes carry source classification confidence and tier. Recall results expose provenance to consumer. Health endpoint reports epistemic state distribution.

7. Retroactive re-evaluation. Competence-change events (graduation transitions, rule accuracy changes) trigger selective logchain replay. The system re-evaluates what it knows when it learns to know better.

Implementation: ~270 lines across 8 files. No new infrastructure. Seb's engineering review required before code — 6 questions in the amendment (schema migration, confidence floor value, bloom filter persistence, ranking weights, reranker context, competence-change triggers).

L0 readiness condition. This amendment is prerequisite for L0 inquiry. If L1 launders uncertainty, L0 inherits false confidence about what the system knows. The contamination problem cannot be investigated on a substrate that exhibits it undetected.

REVIEWED-18 — Observation-Recall Coupling (PENDING-16)

Date: 2026-04-03 Decision: AUTHORIZED Notes:

Proposal to couple the observation path to existing recall capability, replacing blind ingestion with context-aware disposition (novel / reinforcing / noise). Authorized as [PROPOSAL] with the following governance commitments:

1. Logchain immutability preserved. Reinforcement events append linked entries (pointer + similarity score + delta), not mutations. Module stores absorb consolidation. No constitutional change required.

2. Novelty floor invariant. Candidate for Cluster A. Provisional shape: "The system shall not permit its observation disposition to exclude more than [X]% of events from novel classification over any [Y]-day window." Threshold values to be set from empirical data gathered during infant stage -- not by engineering intuition. Draft invariant for jurist review before active-stage graduation.

3. Graduation staging. Three stages authorized (infant → calibration → active). Transition triggers are governance commitments requiring explicit thresholds, to be proposed by executor after infant-stage data collection. No stage transition without steward authorization.

4. Noise audit regime. Audit log retains raw envelope + similarity context + disposition reason + similarity score. 90-day retention aligned to chain pruner. Re-ingestion authorized by steward only. Audit report pushed monthly to steward -- not pulled on request -- to prevent silent filtering gaps from escaping notice. AF-7 external reviewer receives same report once onboarded.

5. Latency budget. Seb to confirm acceptable ingest-path latency for the similarity probe before implementation begins. This is an engineering constraint that shapes the design; do not proceed to code without it.

Activated dead machinery. The proposal connects existing but unwired capability (vector similarity search, CausalEdgeCandidate, compressToAtomicFacts, computeSalience) to the ingestion path. This is architectural completion, not new complexity. Density gate [FIX] (gate ordering in compressToAtomicFacts) authorized separately and independent of this proposal.

REVIEWED-23 — H4 surfacing-to-Seb design-call action (PENDING-21)

Date: 2026-05-14 Decision: AUTHORIZED

Authorize the surfacing action for H4 (hook events pollute recall + logchain accumulation, GH issue #167, priority:high). The PENDING-21 entry's recommended Option B is approved: post a comment on #167 referencing PR #172 + the H2 comment on #165 (batches the audit's three open findings into one Seb-attention window), and surface all three design questions verbatim from the addendum §4 plus the H2/H4 coupling note.

Authorized action shape:

• Comment on CapableMind-ai/betterMemories_app#167
• Asks Seb to address all three questions (prompt filtering at hook / event classification at BMF / recall-triggering policy), or explicitly defer specific ones
• Includes the §6 aggregate-cost framing as a fourth-question-in-effect
• Surfaces the H2/H4 coupling explicitly: H4's ambient cost is pure loss on battery per addendum §6, so resolving H2 toward CPU fallback or default-allow affects H4's cost analysis directly — Seb should see this rather than treating H2 and H4 as independent
• References PR #172 + the H2 comment on #165 to batch attention

Constraint: the comment text itself must be reviewed by the steward before posting. The executor drafts; the steward authorizes the literal text. (Same discipline as REVIEWED-21.)

Escalation path: if Seb's response is "let's talk" or if the comment goes >1 week without engagement, escalate to Option A (sync conversation). Option C (executor pre-drafting a unified design proposal) remains rejected.

No code change in scope. This authorization is for the conversation, not a PR. When Seb's direction is set, the resulting work re-enters the governance loop as a new PENDING.

REVIEWED-22 — Cross-cutting read-path-honest-degradation [PROPOSAL] (PENDING-20)

Date: 2026-05-14 Decision: AUTHORIZED — Option (a) only: PENDING-20 stays as L1 governance entry. L2 elevation work is DEFERRED to post-May 2026 per global CLAUDE.md parked status ("L2 PARKED through end of May 2026. No L2 governance advancement, no new invariant work, no constitutional proposals. L2-adjacent questions arising from L1 work: note, don't pursue.").

Correction (2026-05-14, same session): The original REVIEWED-22 entry recorded Option (c) — both PENDING and L2 elevation. The executor surfaced the L2-PARKED constitutional tension at session end; steward confirmed: "I won't be working on L2 until the end of May — that something was surfaced because of L1 work is both fantastic and coincidental." The surfacing was welcome AS L1-adjacent recognition; the L2 elevation work (cluster declaration, registry entry, contamination-profile exchange) is NOT authorized until end of May 2026. The jurist's framings below are RECORDED for posterity; the work they would shape is held.

Jurist's three governance calls (recorded for post-May pickup):

1. Cluster placement: Not Cluster A. Cluster B (new or existing). The Cluster A case has a real argument but proves too much — it would also pull DN-GOV-05's independent-verifiability test and REVIEWED-19's epistemic integrity into Cluster A. Cluster A is relational posture (system's first-person stance toward the interlocutor — accusative default, asymmetry obligation, precedence of present expression). Read-path observability is epistemically downstream of Cluster A, not peer to it. The common genus with REVIEWED-18 and REVIEWED-19 is conditions under which the system's epistemic behavior can be verified and governed at all: honest ingestion, honest recall, honest degradation. Cluster B is the right cluster. Cluster B may already be implicit (REVIEWED-18, REVIEWED-19 both pending cluster assignment) or need formal constitution; one-sentence steward declaration settles it. Held until post-May.

2. DN-GOV-08 fit: confirmed. The honest-degradation contract sits cleanly on the (i) constituting conditions side, not the (ii) automating recognition side. It specifies structural requirements (the system must expose why it returned empty) without specifying content requirements (what a correct memory result is). The "why empty" breadcrumb tells the operator what happened to the answer-finding machinery, not what the right answer was. Structural distinction, not content judgment. The executor's framing of "stabilizes the conditions" maps cleanly. No rewrite needed. DN-GOV-08 evaluation language to be added to the registry entry per I15/I16/I17 pattern: "This invariant specifies the conditions under which read-path behavior can be observed and governed. It does not specify what correct memory outputs are, or when a recall result is to be trusted. The contract holds open the space for informed steward judgment; it does not automate that judgment."

3. L1_contamination_profile: distinct from Cluster A; requires careful drafting. The contamination risk for a read-path observability invariant is not the same monotonic pressure toward interlocutor satisfaction that Cluster A invariants face. It is the competence-vulnerability paradox from the Observer Problem: the more capable the system becomes at recall, the less visible the failure modes become, because successful recall masks the moments when degradation is silently occurring. Profile candidate: moderate, structural — pressure toward apparent health rather than actual health; manifests as increasing capability masking decreasing observability of degraded paths. Saved as project memory (project-competence-vulnerability-paradox.md) for portability beyond the L2 work. Brief jurist↔steward exchange on the profile language is held until post-May.

Authorized actions THIS session:

• PENDING-20 filed as L1 governance entry — DONE.
• Jurist's three framings recorded above for the post-May pickup.
• Project memory saved for the contamination-shape framing (it is portable to any future read-path work, not just L2).

Held until post-May 2026 (do NOT pursue):

• Steward declaration on Cluster B status.
• Jurist drafting cluster framing note (if needed) and registry entry per I15-I17 pattern.
• Brief jurist↔steward exchange on l1_contamination_profile language.
• Filing of registry entry in capablemind/docs/thinking/David/l2-constitution/amendments/.

Executor's role going forward: maintain PENDING-20 as engineering-visibility record (track Seb's response on H2 #165 and how it informs the cross-cutting); do not produce L2 doctrine; do not initiate cluster declaration or registry-entry drafting; if a future jurist message arrives on this topic before end of May, surface the parked status before responding substantively.

Jurist's three governance calls (authorizing):

1. Cluster placement: Not Cluster A. Cluster B (new or existing). The Cluster A case has a real argument (read-path observability has a conceptual connection to relational posture) but proves too much — it would also pull DN-GOV-05's independent-verifiability test and REVIEWED-19's epistemic integrity into Cluster A. Cluster A is relational posture (system's first-person stance toward the interlocutor — accusative default, asymmetry obligation, precedence of present expression). Read-path observability is epistemically downstream of Cluster A, not peer to it. The common genus with REVIEWED-18 and REVIEWED-19 is conditions under which the system's epistemic behavior can be verified and governed at all: honest ingestion, honest recall, honest degradation. Cluster B is the right cluster. Whether Cluster B is already implicit (REVIEWED-18, REVIEWED-19 both pending cluster assignment) or needs formal constitution is a one-sentence steward declaration. Jurist offered to draft a cluster framing note if needed.

2. DN-GOV-08 fit: confirmed. The honest-degradation contract sits cleanly on the (i) constituting conditions side of the DN-GOV-08 line, not the (ii) automating recognition side. It specifies structural requirements (the system must expose why it returned empty) without specifying content requirements (what a correct memory result is). The "why empty" breadcrumb tells the operator what happened to the answer-finding machinery, not what the right answer was. Structural distinction, not content judgment. The executor's framing of "stabilizes the conditions" maps cleanly. No rewrite needed. DN-GOV-08 evaluation language to be added to the registry entry per I15/I16/I17 pattern: "This invariant specifies the conditions under which read-path behavior can be observed and governed. It does not specify what correct memory outputs are, or when a recall result is to be trusted. The contract holds open the space for informed steward judgment; it does not automate that judgment."

3. L1_contamination_profile: distinct from Cluster A; requires careful drafting. The contamination risk for a read-path observability invariant is not the same monotonic pressure toward interlocutor satisfaction that Cluster A invariants face. It is the competence-vulnerability paradox from the Observer Problem: the more capable the system becomes at recall, the less visible the failure modes become, because successful recall masks the moments when degradation is silently occurring. Profile candidate: moderate, structural — pressure toward apparent health rather than actual health; manifests as increasing capability masking decreasing observability of degraded paths. Different enough from "monotonic" that a brief jurist↔steward exchange is recommended before finalizing the registry entry.

Authorized sequence (jurist's territory; executor does not produce):

1. Steward declares Cluster B status (one sentence). Jurist may draft a framing note to accompany if needed.
2. Jurist drafts registry entry per I15/I16/I17 pattern: statement → sources (architectural/empirical: four H-confirmations, REVIEWED-19, DN-GOV-05 independent-verifiability, bettermemories/CLAUDE.md) → adversarial review → l1_contamination_profile (per Q3 framing) → DN-GOV-08 evaluation (per Q2 language) → residual_risk (deferred pending pilot, per REVIEWED-11 schema decision).
3. Brief jurist↔steward exchange on the l1_contamination_profile language before finalization.
4. Registry entry filed in capablemind/docs/thinking/David/l2-constitution/amendments/ per I15-I17 pattern.

Executor's role going forward: maintain the L1 PENDING-20 entry as the engineering-visibility record (track Seb's response on H2 #165 and how it informs the cross-cutting); do not produce L2 doctrine. When jurist completes the registry entry, executor updates PENDING-20 with cross-reference to the L2 entry.

No code change in scope. No PR. Mechanism work cannot start until Seb chooses a direction for H2 (PENDING-19), and even then would re-enter the governance loop as a fresh PENDING with its own tag.

REVIEWED-21 — H2 surfacing-to-Seb design-call action (PENDING-19)

Date: 2026-05-14 Decision: AUTHORIZED

Authorize the surfacing action for H2 (battery-power suppression silently fails recall, GH issue #165, priority:critical). The PENDING-19 entry's recommended Option B is approved: post a comment on #165 referencing PR #172 (which Seb is about to look at for H3), and surface all four design questions verbatim from the addendum §2.

Authorized action shape:

• Comment on CapableMind-ai/betterMemories_app#165
• Asks Seb to address all four questions (default policy / visible degradation / CPU fallback / query-vs-ingestion asymmetry), or explicitly defer specific ones
• References PR #172 to batch attention while it's high
• Does NOT pre-decide direction; the four questions are policy and Seb's territory

Constraint: the comment text itself must be reviewed by the steward before posting. The executor drafts; the steward authorizes the literal text.

Escalation path: if Seb's response is "let's talk" or if the comment goes >1 week without engagement, escalate to Option A (sync conversation). Option C (executor pre-drafting a unified design proposal) remains rejected — it would have the executor pre-deciding what is properly Seb's call.

No code change in scope. This authorization is for the conversation, not a PR. When Seb's direction is set, the resulting work re-enters the governance loop as a new PENDING with appropriate tag (likely [HARDENING] for any single-direction implementation; possibly [PROPOSAL] if the direction implies architectural change to system_status).

REVIEWED-20 — Fix H3: Temporal stats fallthrough on text queries (PENDING-18)

Date: 2026-05-14 Decision: AUTHORIZED

[HARDENING] proposal authorized after jurist review, with one modification: spec amendment must explicitly bump temporal-module-spec.md from v1.7 → v1.8 as a gap-filling addition, not a clarification. The current behavior is not a misimplementation of a stated rule — the spec (§7) defines query types as enumerated and TemporalStatsQuery (§7.6) requires explicit type: 'temporal_stats'; the spec is silent on missing/unrecognized filter type. The amendment creates a contract that did not exist.

Authorized amendment shape (per jurist):

• (a) filters.type missing/null → {status: 'ok', results: [], total: 0} (honest empty)
• (b) filters.type unrecognized → same return, optional debug note in development mode
• Preserve temporal_stats handler on explicit type: 'temporal_stats' per existing §7.6
• Amendment must articulate WHY: temporal module is type-dispatched by design; a free-text query without type is a malformed query, not a degraded valid query — empty is the honest response, not a stats blob dressed as content.

Authorized sequence:

1. Amendment in capablemind/docs/thinking/David/l1-reliability/h3-temporal-fallthrough-amendment-2026-05-14.md (gap-filling addition, v1.7 → v1.8).
2. Branch fix/h3-temporal-text-query-fallthrough from main.
3. Failing tests reproducing both fallthrough sites red on main.
4. Minimal Option 1 fix (module-level guard at parseTemporalQueryParams).
5. Full test suite + npm run check + npm run lint per BMF CLAUDE.md.
6. PR designed to merge in <30 min of Seb's attention. Body references PENDING-18, REVIEWED-19 (epistemic integrity), REVIEWED-20.

Cross-cutting [PROPOSAL] authorized for parallel filing: read-path-honest-degradation-contract finding from the addendum may be filed as a separate PENDING-19 [PROPOSAL] before or after the H3 PR opens. Naming the pattern is the deliverable; remediation path not assigned. Do not block on it; it does not block the H3 PR.

REVIEWED-17 — ICP-19: Mandated External Review

Date: 2026-03-28 Decision: AUTHORIZED Notes: Closes AF-7 — the corpus's single most consequential gap. Elevates C-R2 from process recommendation to constitutional requirement. Roman strand (provocatio/intercessio) + relational strand (formative contamination detection). Three institutional phases: founding (steward-appointed human reviewer), mature (Founders Circle external panel), deployment (independent reviewer with contractual standing — hard deadline). Not-embedded criterion defined: not sharing L1 substrate, not system interlocutor, unmediated record access, unedited findings. Findings trigger HOLD (not advisory). Three-cycle FM-R4 marker prevents decorative compliance. DN-GOV-08 evaluation passed: holds open the space. Lifts GOV-05 and GOV-06 suspension clauses — bounded self-repair and temporal authorization shift now operative as design commitments. Four open items before first review cycle: Luke's confirmation, findings threshold classification, governed operations scope, Mauss/nexum Phase 2 deferral. L1 contamination: critical, monotonic — formative contamination is the primary threat and external review is the only mechanism operating outside the formative frame.

REVIEWED-24 — PENDING-S1 — Wrap-up §8 output template (pause statement + negative space)

Date: 2026-05-18 Decision: AUTHORIZED (Class A bundle: S1 + S3 + S8) Notes: Steward conversational authorization 2026-05-18 ("I authorize class A"). Implemented same session: ~/.claude/skills/wrap-up/SKILL.md §8 template gains **Pause statement:** and **Decisions deferred (and why):** as named fields, with explicit annotation that the pause is constitutive (Jurist Q3 Harrison-grounded reasoning: the ligature is laid at departure, not discovered at return) and that the negative space is required for the unborn session to know the scope of what was held back. Closes PENDING-S1. First operational test: the /wrap-up at the end of this session.

REVIEWED-25 — PENDING-S3 — Wake-up §3 binary thread validity gate

Date: 2026-05-18 Decision: AUTHORIZED (Class A bundle: S1 + S3 + S8) Notes: Steward conversational authorization 2026-05-18 ("I authorize class A"). Implemented same session: ~/.claude/skills/wake-up/SKILL.md §3 gains an explicit thread validity gate as the first step of synthesis, with binary outcome (confirmed / stale / superseded) and required one-line reason. If stale or superseded, briefing structure reorders to lead with what changed, not with the thread. Moves the staleness check from permission-in-prose to structurally-required gate. Closes PENDING-S3.

REVIEWED-26 — PENDING-S8 — Symmetria pulse lineage anchor + wake-up traversal tools prescribed

Date: 2026-05-18 Decision: AUTHORIZED (Class A bundle: S1 + S3 + S8) Notes: Steward conversational authorization 2026-05-18 ("I authorize class A"). Two related drifts implemented same session:

• Symmetria §6 pulse gains step 0 — re-anchor craft / ethics / character to lineage (now including the τὸ πρόσφορον-includes-time elaboration in §0). Closes the D2 drift (lineage decorative-rather-than-load-bearing in the pulse procedure). Step 5 also gains an explicit cross-reference to §3 self-flag for aligned-without-tension.
• Wake-up §2.b gains a new b.4 substep prescribing mempalace_find_tunnels (when pulling thread crosses project boundaries) and mempalace_kg_timeline (when reconstruction requires knowing when a fact changed). Closes the B6 leverage gap (traversal tools mentioned in constraints but not in procedure).

Closes PENDING-S8.

REVIEWED-S0-via-commit — PENDING-S0 — Prime Directive elaboration

Date: 2026-05-18 Decision: AUTHORIZED-and-CLOSED-by-direct-steward-commit (no separate REVIEWED-N number; recorded here for cluster completeness) Notes: Per Phase 2 of the audit's revised authorization sequence: steward authored the Directive elaboration (τὸ πρόσφορον — what is fitting — includes the time the task requires...); Jurist shape-reviewed the language as drafted + placement (Option 1: both CLAUDE.md and Symmetria §0); steward committed ~/CLAUDE.md directly (line 12, between μέτρον citation and "decision filter" prose) — one-handed with Kai in sling; executor committed ~/.claude/skills/symmetria/SKILL.md §0 (between gloss and §0 header). Constitutional commitment + operational carrier landed in the same moment. Methodology artifacts at ~/_Dev/CapableMind-AI/docs/thinking/David/methodology/ (brief + shape-review + elaboration). Closes PENDING-S0. Class A authorization (REVIEWED-24/25/26) is the operational unfolding of this constitutional commitment in the skills themselves.

REVIEWED-27 — PENDING-26 — .whisper retirement name-lock

Date: 2026-06-04 Decision: AUTHORIZED (jurist ruling, steward relayed) Notes: Role (a) apparatus prose → .apparatus-meta LOCKED. Role (b) qualifier glyphs ✦/⟐ → UNIFICATION with .chamber-mark/.encounter-mark LOCKED — same semantic objects, different layout position; CSS context handles register, not a separate class. Executor's pre-execution gates both passed same evening: positional audit clean (base rule inline-safe); authoring-form census corrected PENDING-26's scope (bare-word-grep inflation — true: 9 template usages + 1 live content file; 15 chamber-v1 legacy files defer to the cluster pass). Migration proceeds as one gated pass: migrate with locked names → compiled-selector grep → render-verify per type per viewport.

REVIEWED-28 — PENDING-25 — Translated renderings in the ARC corpus

Date: 2026-06-05 Decision: AUTHORIZED (jurist Opinion 2026-06-03, steward concurring) — amended by two jurist rulings 2026-06-04/05 (steward-relayed) — IMPLEMENTED AND LIVE 2026-06-05 Notes: Doctrine adopted as ruled: ARC holds works; languages are non-canonical renderings; the work (not any rendering) holds the corpus slot; implementation via W3 (group by work: at build). Two amendments to the original Opinion, both jurist-ruled and steward-relayed:

• Sequencing REVERSED — implementation authorized before Stage G (conditional: decompose complete + render-verified before Stage G opens). Condition met same day.
• translator: refined — required → optional; absence = self-translation (the steward's own renderings carry no credit; Lune credited for Après la réponse and hoped future serious work). Jurist's scope-boundary language (absence-as-signal presupposes site-wide single authorship) integrated verbatim into frontmatter-spec §2.8. Implemented e2866da, deployed, live byte-verified incl. 301: Matthieu Taliesin heteronym decomposed to EN work + fr/es renderings under the work's slot (/slug/fr/, mirroring /v/ — the address says the doctrine); zero-JS Lingua: selector spec'd as apparatus §VII.i; flat-URL migrate-on-touch ratified as class precedent for the 30-piece Jekyll cohort; excludeRenderings replaced the latent dedupe-to-first bug. Both §I.k filters broke dormancy on the real prose (FR 13×U+202F; ES 4 comillas pairs). Closes PENDING-25. Remaining under this doctrine: bio/CV page-renderings extension (machinery currently posts-only); readiness target late August 2026 ahead of Lune's translation stands.

REVIEWED-29 — SEO crawl infrastructure (no PENDING number — jurist-relay authorization, recorded for completeness)

Date: 2026-06-05 Decision: AUTHORIZED (jurist spec-audit pass on the executor's state-verification doc approved execution; steward concurred — Compass listings ruled stable, head-dedup folded in under the jurist's only-if-mechanical criterion) Notes: Executed 6d6f69d, deployed live. robots.txt with NO-disallow posture (executor's mechanism correction to the jurist's default, adopted: disallow blocks crawling not indexing — a disallowed page cannot show its canonical; correct form = canonical + noindex; rationale recorded in spec so disallow cannot be "helpfully" re-added). Sitemap 245 URLs, every one build-verified to resolve (renderings in; chamber-v1 legacy out per stability gate). Shared templates/head.html (duplication resolved; frontispiece pins head_title); siteRoot absolutization of all URL metadata. Latent defect fixed: og:type was never article (×223). /v/ canonical → current version's actual route + noindex. Feed published/updated split per ADR-005. hreflang closes §VII.i's deferred item. Spec records ×4 (apparatus §SEO; versioning ×2; frontmatter §2.4). Open under this entry: CloudFlare prepends managed Content-Signals (ai-train=no; ClaudeBot/GPTBot/CCBot disallowed) — a dashboard default never steward-decided; policy ruling parked, steward's to make.

REVIEWED-30 — PENDING-28 — The imprint register: .imprint class + spec note

Date: 2026-06-06 Decision: AUTHORIZED (jurist opinion, steward relayed) — IMPLEMENTED AND LIVE same day Notes: Doctrinal ground confirmed clean: apparatus-specification.md §VII.g already distinguishes "the piece's locator | the work's signature"; the spec named the register, the CSS did not. The amendment to REVIEWED-27 is a refinement, not a reversal — the lineage runs .whisper → .apparatus-meta (REVIEWED-27) → .apparatus-meta / .imprint by role (this entry). REVIEWED-27 could not adjudicate the imprint distinction because it had not yet been named. Jurist's condition — discharged: type audit preceded the rename. The audit found the imprint was never mono (base .apparatus-meta carries no font-family — Garamond inherited; the italic was incidental inheritance) and .imprint was drafted as its own positively-declared block (body face, normal posture, even-SC Latin name per the frontispiece treatment), not a relabelled copy. The audit also corrected PENDING-28's scope claim by exactly the REVIEWED-27 miss — partials/: three orphaned footer partials (unreferenced, never compiled) deleted under the byte-gate. Executed: e708890, deployed live. Gate: 312/312 pages differ only by the class swap (machine-classified); spec gains §VII.j The Imprint. Verification method: REVIEWED-27's (compiled-selector grep → render-verify), type audit first per condition. Closes PENDING-28.

REVIEWED-31 — PENDING-29 — Sitemap: bio/cv and renderings included

Date: 2026-06-06 Decision: AUTHORIZED (jurist opinion, steward relayed) — IMPLEMENTED AND LIVE same day Notes: Decoupling argument held: enfilade-hidden names a door-strip layout decision; the crawl-exclusion intent was imputed by the accident of bio/cv being its only carriers when the §SEO rule was written. Amends REVIEWED-29's exclusion. Jurist's flags — discharged: (1) Census run against live source, not memory: SIX enfilade-hidden carriers (the four renderings inherited the flag), not two — the +6 arithmetic held regardless; verified in the diff: exactly bio, cv, and four renderings added, nothing removed, 245 → 251. (2) Class precedent recorded at the site.hs rule AND in spec §SEO: enfilade-hidden hereafter means "absent from the door-strip" and nothing more. Crawl exclusion belongs to a frontmatter noindex mechanism — NOT YET IMPLEMENTED (noindexField covers only /v/ prior versions); the gap is recorded in both places so it is not papered over. Executed: 6afce24, deployed live (sitemap 251 URLs verified). Closes PENDING-29.

REVIEWED-32 — PENDING-30 — Banish-JS posture

Date: 2026-06-07 Decision: AUTHORIZED (jurist concurrence, Option 1, with required amendment + observes-nothing specification — both folded into spec text). Landed ARC 7d9bad3. §III.b font-size control = held exception.

REVIEWED-33 — PENDING-31 — Vignette Part II build-time renderer

Date: 2026-06-07 Decision: AUTHORIZED (jurist concurrence, Option 1). Component-2 scoping note required and landed; diagram drops "browser, runtime"; dwell-test gate unchanged. Landed ARC 4b0970b.

REVIEWED-34 — PENDING-32 — Rendering class inheritance

Date: 2026-06-07 Decision: AUTHORIZED (jurist concurrence, conditional — condition found pre-discharged: live §2.8 already defined the fields; jurist read a stale copy, to be relayed). Clause landed ARC 4e89337; build check ships with Stage-G enforcement.

REVIEWED-35 — PENDING-33 — Essay-versioning truth-up + promotion

Date: 2026-06-07 Decision: AUTHORIZED (jurist approval + steward calls: exemplar → The Ethics of the Reply; edition_label dropped from §6 example, field retained). Executed ARC 392d8c8; spec OPERATIVE (built scope).