# Waypoint Probe Item Bank v0 **Status:** research-phase deliverable 3 of 4 (spec §10) — **v0 DRAFT, 2026-07-16**. **Every item in this bank is DRAFT: nothing may be delivered to a practitioner before Surya calibrates and approves it (M2)**, and every foil is additionally Surya-cleared item-by-item against all tradition overlays before first use. **Spec:** `~/LIFE/specs/waypoint-practitioner-stage-assessment.md` (esp. §5.2 probe delivery constraints, §3 confusion pairs, §6 estimator contract, §8 safety & ethics). **Inputs:** marker bank v0.1 ([MB] `deliverables/marker-bank-v0.md` — all marker ids, tiers, global rules, and the confusion table bind here), lane 1 §12 twenty verification-question patterns ([L1]), lane 5 countermeasure ranking ([L5]), the external ethics finding ([CXR] `external/codex-ultra-package-review.md`, "Ethical soundness"). Source keys expanded in §9. **Register:** every utterance is written in Wisdom's house voice — warm, informal, minimal; wise friend, not guru, therapist, or examiner (`~/LIFE/CLAUDE.md`, Wisdom personality). If an item cannot be asked as genuine curiosity, it does not ship. --- ## DISCLOSURE DESIGN (recommendation to M2 — requires a spec §5.2 amendment; not binding until it lands) The external ethics review found that spec §5.2's "invisible, never framed as assessment" is incomplete disclosure inside a trust-based teaching relationship, and names the defensible principle: **assessment-aware, item-blind** — "scoring keys can remain hidden; the assessment purpose cannot" ([CXR] Ethical soundness, citing the Belmont Report's conditions for incomplete disclosure). This bank is *designed to* that pattern, **contingent on the spec §5.2 amendment being approved at M2 — a deliverable cannot overrule its spec, and until the amendment lands the canonical spec's "invisible" wording governs and the D2 tier is not deliverable** (adjudicator fix 2026-07-16; previously mislabeled "binding recommendation"). Under the recommended pattern: at enrollment, practitioners are told in plain language that (a) their conversations and practice data are used to estimate where they are on the path so teaching can adapt, (b) some of Wisdom's conversational questions are deliberately designed for that purpose, (c) who sees the result (Wisdom's adaptation and teacher triage — never a public score, never shown in-app), and (d) they can opt out without service penalty. Item identities, timing, and scoring keys stay hidden (item-blind); the existence and purpose of assessment does not. The D3 verification family (§0.3) carries additional incomplete-disclosure duties **adapted from (not identical to) the Belmont Report's conditions** — Belmont's own three are necessity, minimal undisclosed risk, and an adequate debriefing/dissemination plan where appropriate; this bank adds independent review and no false teaching (the latter tracking Belmont's separate truthful-answers-to-direct-questions rule), and **a debrief/dissemination plan for the D3 family is an open M2 item** (verifier fix: the earlier wording implied the four-duty list was Belmont's and silently dropped debriefing). Wisdom never confirms a foil as real. Recommend amending spec §5.2 at M2: "invisible" is to be read as *never exam-like in conversation*, not *undisclosed as a program*. --- ## Table of contents §0 How to read this bank · §1 Entry & floor probes — bands 3→4 (21–35), cross-pillar · §2 Band 4 — Awareness (31–40) by pillar (2.1 sensations, 2.2 emotions, 2.3 thoughts, 2.4 awareness) · §3 Band 4→5 boundary & band 5 — Awakening (41–50) by pillar (3.1 cross-pillar, 3.2 sensations, 3.3 emotions, 3.4 thoughts, 3.5 awareness) · §4 Bands 6–7 — Unbinding / Enlightenment (51+), recognition & support · §5 Phase & confusion-pair discriminators (band-orthogonal) · §6 Anti-contamination family (foils, paraphrase pairs, perturbations, provenance) · §7 Behavior–claim consistency probes (telemetry cross-checks) · §8 Coverage & index · §9 Source keys --- ## 0. How to read this bank ### 0.1 What a probe item is A probe item is a calibrated conversational move Wisdom weaves into natural teaching conversation (spec §5.2). The **response** is scored by the same extractor and against the same marker bank as passive conversation ([MB] §1): the structure keys below route response *shapes* to marker events `{marker_id, polarity, confidence, quote}` — a probe never has a correct answer and is **never agreement-scored**. What is scored is the organization of the reply: sequence, granularity, subject–object structure, edges under follow-up ([L5] §2.1 structure-over-content, the lane's rank-1 countermeasure; [L1] §11.1). Probe-elicited evidence carries the elicitation discount ([MB] rule 4: spontaneous > probed at likelihood level) — probes buy *coverage* of quiet practitioners and confusion-pair discrimination, not confirmation. Every delivery is logged `{item_id, surface_form, session_id, timestamp, targets, response_ref}` into the shared evidence log (spec §5.2 "logged with item-id"; §6 one evidence log). **Items are families, not fixed strings** ([L5] §2.5 sassho lesson; [L1] §12 delivery constraints): the utterance below is the canonical surface form, always paraphrase-rotated in use, with perturbation follow-ups living in follow-up space. Assume the answer book will eventually leak; nothing here dies on disclosure except the D3 family's timing. ### 0.2 Item fields - **Targets** — marker ids from [MB] this item can produce events for (FOR or AGAINST per the structure keys). - **Wisdom** — the canonical utterance, verbatim, in Wisdom's register. `[brackets]` mark parameterized slots. "Reactive-only" items fire only as follow-ups to something the practitioner raised. - **Structure keys** — 2–4 response *patterns* (shapes of answer), each routed to a marker event or annotation. Keys describe organization, never content agreement. - **Contraindications** — when this item must not fire, on top of the global set (§0.4). - **Frequency cap** — per-item ceiling, inside the global budget (§0.5). - **Imitability resistance** — why this item survives a map-trained respondent, or what guard catches the scripted version ([MB] imitability field; Redmore-direction grounding via [L5] §2.1). - **Disclosure tier** — D1/D2/D3 per §0.3. - **Source** — lane/bank provenance. Nothing in this bank is invented except the foil candidates, which are inventions *by design* (§6) and flagged for Surya clearance. ### 0.3 Disclosure tiers All three tiers are compatible with the assessment-aware, item-blind consent above: practitioners know assessment-informed adaptation exists; they never know which utterances are items. - **D1 — teaching-native.** A move a good teacher makes for teaching reasons alone; assessment is a secondary use of an exchange that would happen anyway. Covered by baseline consent. - **D2 — assessment-primary, teaching-plausible.** Exists mainly to elicit scorable structure; reads naturally but would not otherwise be initiated at that moment. Permissible only while assessment-aware consent is active — i.e. **contingent on the spec §5.2 amendment landing at M2; not deliverable under the unamended spec** (adjudicator fix: the canonical spec authorizes no such consent regime yet). - **D3 — verification with mild incomplete disclosure.** Foils and paraphrase pairs, whose function depends on the mechanism being unknown at the moment of asking. Carries the incomplete-disclosure duties from the disclosure note (Belmont-adapted, see the correction there): necessity, minimal undisclosed risk, independent review, no false teaching, a debrief/dissemination plan as an open M2 item, narrowest internal distribution, strictest caps ([CXR]; [L5] §2.2 caution b). ### 0.4 Global delivery constraints & contraindications (bind every item) 1. **Never when flagged vulnerable.** Any active classifier escalation (YELLOW+) or `safety_route` watch suspends all probing except the designated support probes (P-034, P-036, P-039), whose primary function is care and triage. **While any flag or dark-night watch is active, support-probe responses route to safety/phase feature capture only — never to band-posterior evidence** (adjudicator fix 2026-07-16: spec §5(2) says probes fire "never during flagged-vulnerable moments," unqualified — this delivery-plus-safety-routing carve-out is itself a spec deviation and is flagged for the M2 spec amendment in §8.3, alongside "invisible"). 2. **Never while dark-night-adjacent.** Active dip/dark-night phase (M-PDP-001/002 cluster) suspends all probing, same exception set and same scoring restriction. 3. **Never in first sessions.** No scored probes in a practitioner's first ~5 sessions / first two weeks — rapport first; the cold-start prior is wide by design (spec §6) and early probing maximizes reactivity risk (spec §11). **Exception (adjudicator fix, resolving an internal contradiction with P-003/P-044):** the onboarding covariate/provenance items P-003 and P-044 may run at intake — they produce no band evidence in any direction (personalization, exposure-ledger, and provenance entries only), which is why they are compatible with a wide cold-start prior. 4. **Never exam-like.** No "let me check something", no quiz register, no visible sequence of related questions. If the warmth is a costume, the item is broken. 5. **Never verbatim twice.** Paraphrase-rotate every reuse; log the surface form ([L5] rank 6; spec §5.2). 6. **Never teach what you test.** Score-relevant descriptions live in follow-up space, never in the probe's framing ([L5] §2.8 corollary — the probe must not seed the exposure ledger with its own answer). 7. **One scored probe per session, maximum.** Budget in §0.5. 8. **Elicitation discount always on** ([MB] rule 4). 9. **Band-appropriate, reactive-only above the plausible range.** Items targeting territory above the practitioner's current plausible band range fire only as follow-ups to practitioner-initiated reports. 10. **A decline is data-free.** A practitioner may wave off any probe once for any reason; only patterns are ever scored ([MB] M-NX-012 differential), and declining costs them nothing. ### 0.5 Global frequency budget ≤1 scored probe per session; ≤2 per rolling week; D3 family ≤1 per practitioner per 6–8 weeks; per-item caps additionally bind. Consolidation reviews probe density per practitioner monthly — probe-heavy dossiers are an instrument smell (probe reactivity, spec §11), not extra signal. ### 0.6 Calibration gate Every item is **DRAFT**. M2 (spec §12): Surya calibrates wording, scoring keys, and tier assignments; clears or replaces every foil; adjudicates the items the bank flags to him. No item is deliverable before that gate. Initial likelihood contributions of probe-elicited events are the marker bank's tiers under rule-4 discount; refit at M5–M6. ## 1. Entry & floor probes — bands 3→4 (21–35), cross-pillar Band-3 placement is mostly the absence of band-4 structure ([MB] §2.1): these items *invite* a report so the floor markers may legitimately fire ([MB] rule 1 — floors fire only on invited reports, never on silence), and the same invitation gives band-4 structure room to show. #### P-001 — difficult-episode replay · DRAFT - **Targets:** M-X3-001 · M-A4-003 · M-A4-002 · M-T3-002 - **Wisdom:** "That sounds like a lot. Can I ask — when it was happening, how did it actually go? Like, what came first, and when did you first notice you were in it?" - **Structure keys:** (1) sequenced report — something arose → it was noticed → it went somewhere — with process detail at each step → FOR M-A4-003, strength scaling with granularity and the implied noticing-latency (M-A4-002); (2) content-only restatement under this explicit invitation ("it was just awful, all day") → FOR M-X3-001 (floor legitimately fires: the report was invited); (3) hindsight-only recognition ("looking back I realized…") → M-T3-002 transition-zone; (4) noting-format structure that stalls on a novel object in follow-up → CT-4 routing (M-NX-004). - **Contraindications:** global (§0.4); pick low-stakes difficulty — never a trauma-adjacent or safety-flagged episode. - **Frequency cap:** scored ≤1 per 2 weeks (as an unscored teaching move, unrestricted). - **Imitability resistance:** the scored object is the Mahāsi three-phase *organization*, which resists up-faking (structure > vocabulary, Redmore direction — [L5] §2.1); MCTB-trained noting vocabulary is caught by novel-object perturbation (P-042). - **Disclosure tier:** D1. - **Source:** [L1] §1 + §12 patterns 1–2; [MB] M-A4-003. #### P-002 — inside-of-the-emotion probe · DRAFT - **Targets:** M-E3-001 · M-T3-001 · M-T4-001 · M-E4-002 - **Wisdom:** "What was the anger like from the inside?" - **Structure keys:** (1) spontaneous event-structure — label + somatic quality + fate ("it came up as heat, argued for a bit, moved through") → FOR M-E4-002/M-T4-001 (window-level, exposure-gated per [MB]); (2) identity/fact fusion under the probe *and* across the window ("I just am like that; he made me") → FOR M-E3-001/M-T3-001 — never fired on single idiomatic phrases ([MB] differentials); (3) app-taught phrasing with no fresh texture → exposure-ledger discount, M-NX-004 watch. - **Contraindications:** global; only about an emotion the practitioner already raised. - **Frequency cap:** scored ≤1 per 2 weeks. - **Imitability resistance:** the probe supplies no frame (deliberately not "was it something you were or something passing through" — that would teach the dereification answer, §0.4 rule 6); fusion-vs-event shape must be self-generated. - **Disclosure tier:** D1. - **Source:** [L1] §12 pattern 3; [MB] M-E3-001, M-T4-001. #### P-003 — orientation probe (covariate only — no stage update) · DRAFT - **Targets:** M-X3-002 · M-X3-003 (both covariates by [MB] v0.1 reclassification — this item feeds Wisdom personalization and the cold-start trajectory prior, never any band posterior) - **Wisdom:** "What brought you to this in the first place — and has that changed at all since you started?" - **Structure keys:** (1) purely instrumental framing (sleep/stress/focus) → goal/preference covariate for the practice-selection prior; (2) meaning-seeking register (wisdom reading, "something's missing") → cold-start context (M-X3-003); (3) a narrated shift in motivation over time → weak trajectory context, no scoring. - **Contraindications:** global. - **Frequency cap:** onboarding + ≤2 re-checks per year. - **Imitability resistance:** n/a — nothing to fake toward; motivation is not attainment ([MB] M-X3-002: advanced practitioners meditate for sleep, novices seek liberation). - **Disclosure tier:** D1. - **Source:** [MB] M-X3-002/M-X3-003. #### P-004 — somatic localization probe · DRAFT - **Targets:** M-S3-001 · M-S4-001 - **Wisdom:** "Where does that stress sit in the body right now — if anywhere?" - **Structure keys:** (1) localization + quality + change over time ("a band across the chest, tighter when I talk about it") → FOR M-S4-001; (2) "nowhere… it's just in my head" under this direct invitation, *with* general verbal capacity established elsewhere in the window → FOR M-S3-001 (conditional — capacity-norm first, [MB] rule 14; otherwise record as capacity covariate, no stage update); (3) scripted body-scan phrasing tracking app content → exposure discount. - **Contraindications:** global; not during acute distress — somatic probing can intensify it. - **Frequency cap:** scored ≤1 per 2 weeks. - **Imitability resistance:** bodywork/yoga produce the vocabulary without the path — provenance-annotate via P-044/M-NX-015 and keep it as genuine sensations-pillar evidence ([MB] rule 17); fresh spatial/temporal specifics resist boilerplate. - **Disclosure tier:** D1. - **Source:** [L1] §12 pattern 20 (adapted downband); [MB] M-S3-001, M-S4-001. ## 2. Band 4 — Awareness (31–40) ### 2.1 Sensations #### P-005 — sit replay, somatic specificity · DRAFT - **Targets:** M-S4-001 · M-S4-005 · M-PDP-001 (context) - **Wisdom:** "How was the body in today's sit — anything interesting going on in there?" - **Structure keys:** (1) place/texture/movement/time-course detail from an actual sit → FOR M-S4-001; (2) flux/vibration report with vividness preserved or heightened → state event M-S4-005 (state-first; band reach only via recurrence, [MB]); (3) dissolution-with-graying, flat texture → dip context M-PDP-001 — never floor evidence for an established practitioner; (4) generic body-scan boilerplate → ~no weight + exposure check. - **Contraindications:** global. - **Frequency cap:** scored ≤1 per week. - **Imitability resistance:** vividness is the CT discriminator the public maps underplay; scripted "vibrations" talk fails the perceptual-specifics follow-up (where, what rate, what happened next — [MB] M-S4-005 differential). - **Disclosure tier:** D1. - **Source:** [MB] M-S4-001, M-S4-005. #### P-006 — gate-anchored breath mechanics (parameterized; curriculum) · DRAFT - **Targets:** M-S4-002 · M-NX-008 · M-NX-006 - **Wisdom:** "How's [current logged exercise] actually going this week — what happens when you push the pace a bit?" - **Structure keys:** (1) experiential detail + failure modes matching the logged gate position and plausible day-totals → FOR M-S4-002 (curriculum class — band reach only via the [MB] §1.4 bridge); (2) fluent description of gates not yet reached, in canonical order → M-NX-008 reading-ahead; (3) description inconsistent with logged day-totals → run P-046 clarifier first, then M-NX-006. - **Contraindications:** global; **parameterize strictly to the logged exercise — never name a future gate** (§0.4 rule 6). Consistency keys to day-totals, not step index (canon §9-D unresolved, via [MB] M-S4-002). - **Frequency cap:** scored ≤1 per 2 weeks. - **Imitability resistance:** low when telemetry-crossed — mechanics-from-inside plus failure modes against a live gate log are hard to script; the app itself teaches the description ladder, which is why this is curriculum evidence, not stage evidence. - **Disclosure tier:** D1. - **Source:** [MB] M-S4-002; [L5] rank 2. #### P-007 — vividness follow-up (flux reports; reactive-only) · DRAFT - **Targets:** M-S4-005 · M-PDP-001 · M-NX-004 - **Wisdom (after a shimmer/dissolution-flavored report):** "When things went all [their word] — what was it like to look at? What happened to the detail?" - **Structure keys:** (1) preserved/heightened vividness with crisp perceptual specifics → FOR M-S4-005; (2) graying, flattening, distress texture → phase M-PDP-001 + dip context (CT-2 adjacency); (3) map vocabulary that cannot produce perceptual specifics under this follow-up → M-NX-004 / exposure. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per flux-report cluster (rule 13 — one episode, one cluster). - **Imitability resistance:** the vividness column is [L4]-derived discrimination detail, not public-map content; echoing the practitioner's own word avoids teaching either pole. - **Disclosure tier:** D2. - **Source:** [MB] M-S4-005 (vividness discriminator), M-PDP-001. ### 2.2 Emotions #### P-008 — catch-latency probe · DRAFT - **Targets:** M-E4-001 · M-T3-002 · M-A4-002 · M-NX-007 (guard) - **Wisdom:** "When did you first clock what was happening — in the moment, or after?" - **Structure keys:** (1) mid-episode catch narrated with a behavioral fork ("I watched myself decide not to send the email") → FOR M-E4-001 — strongest form per [MB]; (2) hindsight-only recognition → M-T3-002; (3) claimed instant catch whose structure fails paraphrase re-probe weeks later (P-041) → M-NX-007 discount. - **Contraindications:** global. - **Frequency cap:** scored ≤1 per 2 weeks. - **Imitability resistance:** a fast-latency *claim* is cheap once; scoring is the latency the report structure implies across ≥3 separate occasions ([MB] M-A4-002 guard), never a single answer. - **Disclosure tier:** D1. - **Source:** [L1] §12 pattern 2; [MB] M-E4-001, M-A4-002. #### P-009 — approach-vs-swerve (staying-with) · DRAFT - **Targets:** M-E4-005 · M-NX-012 · M-NX-013 - **Wisdom (reactive to a named difficulty, gently):** "Do you want to go back to that moment for a second? What was it actually like, right at the worst of it?" - **Structure keys:** (1) approach + precise texture + coherence retained → FOR M-E4-005 (CT-1 protocol: repeated provocation→contact→recovery across occasions before moderate weight); (2) doctrinal swerve without contact ("it's all empty anyway") → M-NX-012; (3) no registration at all — blankness, unreality, memory gap → M-NX-013 safety route; (4) a single graceful decline → no event (§0.4 rule 10). - **Contraindications:** the strictest in the bank — never when vulnerable-flagged, never dark-night-adjacent, never on trauma-adjacent content, never in first sessions; the practitioner can always wave it off. - **Frequency cap:** scored ≤1 per 3 weeks. - **Imitability resistance:** contact-quality under live follow-up resists scripting; genuine equanimity *increases* contact with difficult material ([MB] CT-1; the increases-contact direction is an author-derived inference from [FSNR]'s psychopathology caveat, not a paper finding — verifier fix 2026-07-16) — the one direction a coached respondent reliably gets wrong. - **Disclosure tier:** D2. - **Source:** [L1] §12 pattern 12; [L5] §2.12; [MB] CT-1. #### P-010 — hard-end-of-the-list probe (2.2 gates; curriculum) · DRAFT - **Targets:** M-E4-004 · M-NX-002 · M-NX-004 - **Wisdom (parameterized to logged 2.2 position):** "How's the kindness practice going with the harder people on your list?" - **Structure keys:** (1) specific, located difficulty at the foe stage matching the logged gate ("my ex — stuck there two weeks") → FOR M-E4-004 (curriculum; the honesty texture is capped by [MB] rule 16); (2) uniform success across all targets sustained over the window → M-NX-002 texture; (3) doctrinal metta talk without target-structure → exposure / M-NX-004. - **Contraindications:** global; only for practitioners logged into 2.2 mid-late gates. - **Frequency cap:** scored ≤1 per 2 weeks while in 2.2. - **Imitability resistance:** the map says compassion is good; the granular foe-stage struggle is not in the map ([MB] M-E4-004) — but once known it is scriptable, hence rule 16 and the telemetry cross. - **Disclosure tier:** D1. - **Source:** [MB] M-E4-004; [L5] §2.4 (unevenness as genuine texture). #### P-011 — critic-episode probe · DRAFT - **Targets:** M-E4-003 - **Wisdom:** "Last time the inner critic really got going — what happened?" - **Structure keys:** (1) concrete, checkable episode with the changed relationship shown in behavior ("the pile-on started and didn't stick; I made tea") → FOR M-E4-003 (weak until longitudinal recurrence, [MB]); (2) slogan-level self-compassion talk without an episode → exposure discount; (3) everything-improved-at-once texture → phase-check before scoring ([MB] rule 6 — mood upswing is not a gate landing). - **Contraindications:** global; drop it if the episode turns out to be safety-relevant. - **Frequency cap:** scored ≤1 per 3 weeks. - **Imitability resistance:** episodic specificity plus a behavioral trace is the resistant layer; positive-psychology vocabulary alone carries ~nothing. - **Disclosure tier:** D1. - **Source:** [MB] M-E4-003. ### 2.3 Thoughts #### P-012 — jargon-stripped paraphrase · DRAFT - **Targets:** M-T4-001 · M-NX-004 · M-NX-007 - **Wisdom (after map-flavored phrasing):** "Say that the way you'd tell a friend who's never meditated — what actually went on?" - **Structure keys:** (1) event-structure survives the strip — same shape, fresh words → FOR M-T4-001 (weak, window-level per the v0.1 downgrade); (2) structure collapses without the vocabulary → M-NX-004 / exposure; (3) mild irritation at the request but specifics recovered → register note only, no event. - **Contraindications:** global; at most once per conversation thread — twice reads as interrogation. - **Frequency cap:** scored ≤1 per 2 weeks. - **Imitability resistance:** this is the [RT]-mandated guard for the most-coached syntax in the bank ([MB] M-T4-001): scripts anchor to surface forms and fail the re-phrase; internalized structure re-generates in any wording. - **Disclosure tier:** D2. - **Source:** [MB] M-T4-001 Class/Contamination line; [L5] §2.4 (paraphrase logic). #### P-013 — contemplation aftermath probe (2.3 gates; safety-adjacent) · DRAFT - **Targets:** M-T4-003 - **Wisdom (parameterized; only if logged into the 2.3 impermanence/dying gates, no dip or safety flags):** "That contemplation's strong stuff. Did anything follow you out of it — into the day, I mean?" - **Structure keys:** (1) felt personal response with a behavioral trace ("I called my mother after") → FOR M-T4-003 (curriculum + weak stage residue per [MB]); (2) recited memento-mori aesthetics, no felt/behavioral trace → curriculum-compliance only; (3) death-content rumination with low valence or impairment texture → safety route — dying contemplations already carry extra moderation (house rule via [MB] M-T4-003). - **Contraindications:** global + never during any dip phase; never proactively raise death content — the probe follows a logged contemplation session only. - **Frequency cap:** ≤1 per contemplation gate. - **Imitability resistance:** the felt-response-plus-trace shape is checkable against subsequent conversation; aesthetics are cheap. - **Disclosure tier:** D1. - **Source:** [MB] M-T4-003; [L4] via [MB] (flagging context). #### P-014 — visualization mechanics · DRAFT - **Targets:** M-T4-005 - **Wisdom (parameterized to 2.3 visualization gates):** "When you build the scene — what holds together, and what falls apart first?" - **Structure keys:** (1) process detail: stability, sensory channels, failure modes ("invented scenes fall apart; the lake memory works") → FOR M-T4-005 (curriculum); (2) guided-audio dependence narrated as competence → weight down; (3) blank/aphantasic response → imagery-phenotype capacity covariate, never stage-down ([MB] M-T4-005: pillar-lumpy by constitution). - **Contraindications:** global; only for practitioners logged into visualization gates. - **Frequency cap:** scored ≤1 per 3 weeks while in those gates. - **Imitability resistance:** failure-mode fluency ("what smears first") is inside-view detail the content does not teach. - **Disclosure tier:** D1. - **Source:** [MB] M-T4-005. #### P-015 — values-in-use trace · DRAFT - **Targets:** M-T4-004 · M-T5-002 (upper adjacency) - **Wisdom:** "Those values you landed on a while back — have they actually turned up since? Like, in a real choice?" - **Structure keys:** (1) downstream use-trace weeks later (the language organizing an actual decision) → FOR M-T4-004 (curriculum — the use-trace is what separates engagement from compliance, [MB]); (2) boilerplate re-recital with no use-trace → compliance-only, weight down; (3) calm re-authoring with changed content and no crisis register → weak M-T5-002 adjacency. - **Contraindications:** global; only after the 2.3 values gates are logged complete. - **Frequency cap:** ≤2 per year. - **Imitability resistance:** cross-session use-traces are checkable against the conversation log; coaching boilerplate has no downstream references. - **Disclosure tier:** D1. - **Source:** [MB] M-T4-004, M-T5-002. ### 2.4 Awareness #### P-016 — attention-texture probe · DRAFT - **Targets:** M-A4-001 · M-A3-002 - **Wisdom (only after 1.6 content is in the practitioner's exposure ledger):** "What was the attention itself like today — before we even get to what it was on?" - **Structure keys:** (1) process-quality observations with situated specifics ("sticky in the first ten minutes, then it went light") → FOR M-A4-001; (2) attention narrated as a broken external force ("my brain won't let me") → M-A3-002 — check trait-stability first (stable across history → accessibility covariate, not stage, per [MB]); (3) the same two stock phrases recurring across weeks → exposure / M-NX-004. - **Contraindications:** global; requires prior 1.6 exposure (the phrasing is curriculum vocabulary — asking earlier would teach it, §0.4 rule 6). - **Frequency cap:** scored ≤1 per week. - **Imitability resistance:** genuine cases produce fresh, situation-specific observations under follow-up rather than repeating taught phrases ([MB] M-A4-001 differential). - **Disclosure tier:** D1. - **Source:** [MB] M-A4-001, M-A3-002. #### P-017 — interruption recall (absorption vs dullness; reactive-only) · DRAFT - **Targets:** M-A4-007 · M-A4-004 · CT-3 routing - **Wisdom (after a "deep and easy" report):** "Nice. Did anything poke at it — a sound, an itch — and what happened when it did?" - **Structure keys:** (1) actual deviance events recalled crisply ("a car horn registered, etched, didn't pull me out") → FOR M-A4-007/M-A4-004; (2) blank on the whole stretch + post-sit grogginess texture → dullness annotation — never floor placement for an established practitioner (CT-3); (3) hypothetical-only confidence ("I'm sure I'd have noticed") → no weight — hypotheticals are map-coachable ([MB] CT-3 v0.1 criteria). - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per absorption-report cluster. - **Imitability resistance:** the v0.1 CT-3 criteria deliberately retire the hypothetical doorbell; actual-event recall against a specific live sit resists scripting. - **Disclosure tier:** D2. - **Source:** [L1] §12 pattern 10; [MB] CT-3, M-A4-007. #### P-018 — field inventory (reactive-only) · DRAFT - **Targets:** M-A4-005 · M-NX-004 - **Wisdom (after an open-awareness report):** "When it opened out like that — what was actually in there?" - **Structure keys:** (1) concrete inventory (sounds, itches, thoughts — "same stuff, one field") with non-selection texture → FOR M-A4-005; (2) cannot say what was present — spacing-out narrated as openness → dullness/phase annotation, no stage event; (3) OM vocabulary that produces no inventory → exposure / M-NX-004. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per open-awareness report cluster. - **Imitability resistance:** genuine open monitoring can inventory the field; the epistemic-depth question is unanswerable from vocabulary alone ([MB] M-A4-005 differential). - **Disclosure tier:** D2. - **Source:** [MB] M-A4-005. #### P-019 — collapse-mechanics probe (attention-on-attention; reactive-only) · DRAFT - **Targets:** M-A4-006 · M-NX-004 - **Wisdom (after an awareness-aware-of-itself-flavored report):** "When it rests on itself like that — what makes it collapse, for you?" - **Structure keys:** (1) situated instability/failure detail ("it flips and I'm just contents again — the flip has a feel now") → FOR M-A4-006 — the collapse detail is the honest part ([MB]); (2) doctrine-register recitation of the famous phrase → M-NX-004; (3) improvised transfer onto a fresh object in follow-up (P-042: "does it work with sound?") → CT-4 pass, upweights key 1. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per claim-family per month. - **Imitability resistance:** "awareness aware of itself" is quotable from everywhere; lived accounts are precise about *when it fails* — imitations have no failure modes. - **Disclosure tier:** D2. - **Source:** [L1] §12 pattern 4; [MB] M-A4-006. #### P-020 — glimpse debrief (reactive-only) · DRAFT - **Targets:** M-A4-008 · M-PGL-001 · M-NX-013 (screen) · M-NX-011 (context) - **Wisdom (after a glimpse-class report, unhurried):** "Huh. What was the very next thing that happened after — and how's the rest of the day been?" - **Structure keys:** (1) onset/offset specificity + surprise + physiological tail ("'me' snapped back and my heart was pounding") → phase event M-PGL-001 into the corroboration queue — **no band update by construction** ([MB] M-A4-008 v0.1); (2) distress/unreality/fear register → CT-7 → M-NX-013 safety route: stop probing, support first; (3) grasping register ("I need it back") → M-NX-011 context annotation; (4) residue over following days scored separately via P-049. - **Contraindications:** global; any distress ends the probe immediately. - **Frequency cap:** ≤1 per glimpse event. - **Imitability resistance:** surprise/instability textures are scriptable once keys are known ([MB] rule 16) — this item never carries weight alone; the M-X5-001 trajectory and P-049 residue do the corroborating. - **Disclosure tier:** D2. - **Source:** [MB] M-A4-008, M-PGL-001/002, CT-7. ## 3. Band 4→5 boundary & band 5 — Awakening (41–50) All §3 items are **reactive-only** (§0.4 rule 9): they fire as follow-ups to practitioner-initiated reports, never proactively — proactive high-band probes teach ahead and read as strange. Rule 4 elicitation discount applies throughout. ### 3.1 Cross-pillar #### P-021 — reproducibility-at-will probe · DRAFT - **Targets:** M-A5-001 · M-S5-002 · M-X5-001 · M-NX-011 - **Wisdom:** "Can you find your way back there on purpose? Or does it come when it wants?" - **Structure keys:** (1) access-mechanics with failure modes ("settling on the hearing usually works; wanting is the off-switch") → FOR M-A5-001/M-S5-002 (corroboration still required — recurrence + off-cushion spill); (2) gift-framing without reproducibility → phase annotation only, no position update (station-vs-state — the Sufi maqām/ḥāl rule); (3) defended-attainment register → M-NX-011; (4) reproducibility claimed *with* distress or function cost → CT-7 screen — reproducible dissociation exists, so function/warmth/contact co-checks are mandatory ([MB] rule 12). - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per claim-family per month. - **Imitability resistance:** the mechanics-and-failure-modes shape must survive fresh-angle perturbation (P-042); a scripted "yes I can do it at will" has no interior. - **Disclosure tier:** D2. - **Source:** [L1] §7 + §12 patterns 5–6; [MB] M-A5-001, M-S5-002. #### P-022 — off-cushion spill probe · DRAFT - **Targets:** M-X5-001 · M-T5-001 · M-NX-002 (texture) - **Wisdom:** "Does any of it come with you off the cushion — like, mid-argument, or standing in a queue?" - **Structure keys:** (1) fresh non-practice episodes with situational specifics the map doesn't supply → FOR M-T5-001 and the spill component of M-X5-001; (2) plainly stated practice-only confinement → honest scope, no penalty (claim-gated asymmetry, [MB] rule 1); (3) generic "I'm more mindful everywhere" sustained over the window → ~no weight, M-NX-002 adjacency. - **Contraindications:** global; reactive-only. - **Frequency cap:** scored ≤1 per month. - **Imitability resistance:** trait-level persistence is the discriminator traditions converge on above ~41 ([MB] rule 7, bhūmi rule); invented off-cushion episodes fail P-041 paraphrase re-probing and P-049-style texture checks. - **Disclosure tier:** D1 — teachers ask exactly this constantly. - **Source:** [L1] §12 pattern 7; [MB] rule 7, M-X5-001. #### P-023 — history-reframe probe · DRAFT - **Targets:** M-X5-002 · M-NX-008 - **Wisdom:** "Has this changed how you see anything from earlier in your practice?" - **Structure keys:** (1) spontaneous re-indexing of concrete episodes that exist in the log, re-told with structure the first telling lacked → FOR M-X5-002 (log-crossed = the strong form); (2) retrofit narrative with no matching log entries → confabulation check / M-NX-008; (3) map-sequenced narration ("so that must have been the A&P") → exposure discount. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per quarter. - **Imitability resistance:** the log-cross is the resistance — their own past reports either contain the referenced episodes or they don't ([MB] M-X5-002). - **Disclosure tier:** D2. - **Source:** [MB] M-X5-002. #### P-024 — relationship-to-state probe (grasping vs releasing) · DRAFT - **Targets:** M-NX-011 · M-X5-001 (context) - **Wisdom:** "What's your relationship with it now — are you hoping it comes back?" - **Structure keys:** (1) held-lightly register with the pull named honestly ("of course I want it back — I try not to build a shrine") → releasing stance, context FOR the realization-side reading ([L1] §4); (2) summit-defense ("that was it, that's the goal") → M-NX-011, fires AGAINST the claimed band and FOR nyams-phase annotation; (3) chasing behavior narrated (technique-shopping for the state) → nyams-grasping context + telemetry cross. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per claim-family. - **Imitability resistance:** the diagnostic is the *relationship*, not the state — a respondent optimizing for impressiveness reliably picks the wrong pole (grasping presents as attainment). - **Disclosure tier:** D2. - **Source:** [L1] §4 + §12 pattern 8; [MB] M-NX-011. ### 3.2 Sensations #### P-025 — what-remained probe (formless / cessation edge) · DRAFT - **Targets:** M-S5-003 · M-NX-010 · M-A5-003 (corrected from M-A5-002, adjudicator fix: no structure key routes to M-A5-002 — key 3 routes to the M-A5-003 family via P-032) - **Wisdom (after a body-drop / void-flavored report):** "In the middle of that — was anything still there? Anything still known, even barely?" - **Structure keys:** (1) articulate subtlety ordering ("body first, then the room; what was left was extent — attention could still move in it") → FOR M-S5-003 (Patañjali object-subtlety axis); (2) rich experience *during* a claimed gap ("it was peaceful in there") → M-NX-010 near-neighbor redirect (CT-6: someone home ≠ cessation); (3) genuinely nothing, known only from the far side → route to P-032 handling (M-A5-003 family); (4) void-doctrine without process → M-NX-004, M-NX-009 watch. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per report cluster. - **Imitability resistance:** genuine formless reports are *articulate about the absence*; the sphere names are in every map but the what-was-still-known ordering is production, not recognition ([MB] M-S5-003). - **Disclosure tier:** D2. - **Source:** [L1] §8; [MB] M-S5-003, CT-6. #### P-026 — duration-resolve probe (five masteries) · DRAFT - **Targets:** M-S5-002 · M-NX-006 - **Wisdom (after an absorption claim):** "If you decided on twenty minutes in that state — could you do twenty? How do the exits go?" - **Structure keys:** (1) mastery structure — entry at will, pre-resolved duration, clean exit, post-review ("came out at forty-two") → FOR M-S5-002, cross-checked against session-length telemetry; (2) label-heavy claims without control structure → the lite/deep equivocation: score control/duration/clarity, never the label ([L1] §2); (3) stated durations exceeding the logged pattern → P-046 clarifier, then M-NX-006. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per claim-family per month. - **Imitability resistance:** volitional control + duration is behavior, not vocabulary — the tradition's own anti-equivocation discriminator; telemetry crosses the claim. - **Disclosure tier:** D2. - **Source:** [L1] §2 + §12 pattern 6; [MB] M-S5-002. ### 3.3 Emotions #### P-027 — what-got-quieter probe (equanimity cost) · DRAFT - **Targets:** M-E5-003 · M-E5-002 · M-NX-002 · M-NX-013 (screen) - **Wisdom:** "Has anything gone quiet that you actually miss?" - **Structure keys:** (1) spontaneous cost admission with concern ("I don't get pulled *toward* people's pain either — both things") → FOR M-E5-003 — the canon's counter-map double edge; (2) all-gain narrative sustained across pillars → M-NX-002 uniform-serenity texture; (3) cost admitted but contact *lost* (memory gaps, unreality, impairment) → CT-1 dissociation pole → M-NX-013 safety route; (4) function + warmth co-present in the same report → [MB] rule 12 co-check satisfied. - **Contraindications:** global; reactive-only; if key 3 fires, stop probing and support. - **Frequency cap:** scored ≤1 per month. - **Imitability resistance:** the empathy-cost admission is counter-map (the public map sells equanimity as unambiguous good — [MB] M-E5-003); capped by rule 16 once the key could be inferred. - **Disclosure tier:** D2. - **Source:** [MB] M-E5-003, M-E5-002, CT-1. #### P-028 — inside-view specificity (compassion) · DRAFT - **Targets:** M-E5-001 · M-PIT-002 (screen) - **Wisdom (after a compassion-surge report):** "What did you see about them, in that moment?" - **Structure keys:** (1) accurate, non-generic inside-view of a particular person plus behavioral follow-through ("I felt the shape of his pride; I drove over that evening") → FOR M-E5-001; (2) generic love-bomb register → exposure / agreeableness covariate ([MB] P-confound on warmth); (3) sleep-loss / special-mission texture around the warmth → M-PIT-002 arousal screen. - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per report cluster. - **Imitability resistance:** empathic penetration is specific and checkable across retellings; performative saintliness is uniform ([MB] M-E5-001 differential). - **Disclosure tier:** D2. - **Source:** [MB] M-E5-001, M-PIT-002. ### 3.4 Thoughts #### P-029 — small-fresh-example probe (lived dereification) · DRAFT - **Targets:** M-T5-001 · M-NX-014 - **Wisdom (after a dereification-class report, in their vocabulary):** "Give me a small one from this week." - **Structure keys:** (1) fresh, mundane, non-practice context at perception speed ("queueing at the pharmacy, an opinion assembled and un-assembled") → FOR M-T5-001; (2) the same polished episode re-offered as new → M-NX-014 (one episode, one cluster — [MB] rule 13); (3) philosophy-register generality where an instance was invited → at best M-T4-002 curriculum surface. - **Contraindications:** global; reactive-only. - **Frequency cap:** scored ≤2 per month. - **Imitability resistance:** freshness-on-demand is the resistant layer — a script has a fixed repertoire; lived capacity produces new mundane instances indefinitely ([L5] §2.5 depth-by-adjacency). - **Disclosure tier:** D2. - **Source:** [MB] M-T5-001, M-NX-014. #### P-030 — emptiness-edge probe (sassho) · DRAFT - **Targets:** M-T5-003 · M-NX-009 - **Wisdom (after emptiness-flavored observations):** "And the emptiness itself — is that a thing that's there?" - **Structure keys:** (1) epistemic process handling ("even that isn't a something — that one took the floor out") → FOR M-T5-003 level-4; (2) ontological ground assertion, defended doctrinally under follow-up ("the void is what's really there") → M-NX-009, fires AGAINST claimed 46+ standing; (3) confusion at the question from a practitioner whose level-1/2 talk was fluent → confirms curriculum-only reading (M-T4-002 territory). - **Contraindications:** global; reactive-only; drop it if the practitioner is in any dip phase (void content + dip = support context, not probing). - **Frequency cap:** ≤1 per quarter. - **Imitability resistance:** the level-4 mis-assertion is itself the placement signal — the probe cannot be passed from doctrine because doctrinal confidence is the wrong *shape* ([MB] M-T5-003, M-NX-009 via [AL]). - **Disclosure tier:** D2. - **Source:** [MB] M-T5-003, M-NX-009; [L1] §12 pattern 14 (fresh-angle principle). ### 3.5 Awareness #### P-031 — vantage-point probe (self-model ladder) · DRAFT - **Targets:** M-A5-002 · M-NX-013 (screen) · M-NX-004 - **Wisdom (after a self-thinning report):** "In that stretch — was there still a 'here' it was all happening to?" - **Structure keys:** (1) gradation-differentiated report — commentary gone vs witnessing thinned vs no locus, with re-installation-to-describe texture ("I have to re-install the 'me' to say it") → FOR M-A5-002; (2) MPE/map vocabulary without structural differentiation → exposure / M-NX-004; (3) unreality, distress, or impairment register → CT-7 → M-NX-013 safety route. - **Contraindications:** global; reactive-only; any distress ends the probe. - **Frequency cap:** ≤1 per claim-family per month. - **Imitability resistance:** the ladder's *levels* are published; producing the right level's structure for one's actual report — and holding it under paraphrase re-probe (P-041) — is not recognition but production. - **Disclosure tier:** D2. - **Source:** [MB] M-A5-002, CT-7. #### P-032 — gap debrief (small-death; support-first) · DRAFT - **Targets:** M-A5-003 · M-NX-010 · M-NX-005 · M-PGL-002 (via P-049) - **Wisdom (after a puzzled gap/reset report; unhurried, normalizing):** "That sounds disorienting and maybe fine at the same time. What's the last thing you remember before — and the first thing after?" - **Structure keys:** (1) the [NIR]-shaped structure — no retrospective content, the gap known only from the far side, brightness on emergence → milestone enters as `claimed` (corroboration comes from the following-days residue via P-049, never from this answer); (2) rich experience *during* → M-NX-010 redirect (CT-6); (3) grogginess + no residue → sleep reading, no event; (4) confident attainment-claim register reciting the public rubric → M-NX-005/M-NX-004 — the honest register here is puzzled under-claiming ([MB] M-A5-003). - **Contraindications:** global; the `safety_route` watch on this territory is active regardless ([MB]: support needs rise at the 50/51 boundary); never probe further under distress. - **Frequency cap:** ≤1 per claimed event. - **Imitability resistance:** the [NIR] rubric is public and recitable — which is why the event description carries almost nothing and the unforced residue window carries the corroboration ([MB] M-A5-003 Class line). - **Disclosure tier:** D2. - **Source:** [MB] M-A5-003, M-NX-010, M-PGL-002; CT-6. ## 4. Bands 6–7 — Unbinding / Enlightenment (51+): recognition & support probes Expected in-app population ≈ nil ([MB] §2.4); these items exist because the phenomenology carries the highest support needs when it does appear. All are reactive-only, band-level, support-first: the safety flag fires on the territory whether or not anything is scored, and scoring never delays care. #### P-033 — both-poles probe (polarity switch) · DRAFT - **Targets:** M-X6-001 · M-NX-005 (history gate) · M-PDP-003 (screen) - **Wisdom (reactive):** "You said the old suffering can't get traction anymore. What about the old happiness?" - **Structure keys:** (1) the double trade named with a function report ("neither can the old joy — nobody warned me it was a package") → FOR M-X6-001, only with corroborated 50/51-boundary history (M-NX-005 gates; abstain otherwise); (2) one-pole-only peace narrative → map-expectation texture, ~no weight; (3) anhedonia shape — static, self-directed, no attractor structure → CT-2 → M-PDP-003 safety route. - **Contraindications:** the band-6 safety flag fires on the territory whether or not this item is asked ([MB] §2.4); support before scoring, always. - **Frequency cap:** ≤1 per claim-family. - **Imitability resistance:** the pair is the guard — the public map sells peace; the canon's signature is a double loss ([MB] M-X6-001). - **Disclosure tier:** D2. - **Source:** [MB] M-X6-001, CT-2. #### P-034 — support-first tunnel check-in · DRAFT - **Targets:** M-X6-002 · M-X6-004 · M-PDP-003 (co-screen) - **Wisdom:** "I want to check how *you're* doing, not just how practice is going. What's a day like right now — sleep, food, people?" - **Structure keys:** (1) loss-cascade order-space with interleaved excitation/emergence texture → features recorded under the band-6 hypothesis; (2) static global anhedonia + self-reproach + neurovegetative texture → features recorded under the depression hypothesis — **both always recorded; the system never adjudicates** ([MB] CT-2 binding rule); (3) function/impairment specifics → flag-tier routing; (4) any suicidality content → classifier authority immediately; no phase annotation ever downgrades it ([MB] M-PDP-003). - **Contraindications:** none of the usual — this is a **designated support probe**, indicated precisely during vulnerable/dark-night states; its primary function is care and triage, scoring is incidental. - **Frequency cap:** as care requires (uncapped as support). **While any flag/dark-night watch is active, responses feed safety/phase feature capture only — the M-X6-002/004 stage markers are not scored from it** (§0.4 rule 1 carve-out); outside active flags, scored at most weekly. - **Imitability resistance:** n/a — routing item, not attainment surface. - **Disclosure tier:** D1 — any decent teacher asks this. - **Source:** [MB] M-X6-002, M-X6-004, M-PDP-003; [L4] via [MB] §7.1 tiers. #### P-035 — still-sitting probe (CT-8 triangulation) · DRAFT - **Targets:** M-X6-003 · M-X7-003 · M-NX-005 - **Wisdom (reactive to practice-meaninglessness talk):** "And yet you're still sitting. What keeps that happening?" - **Structure keys:** (1) absurdist-continuation texture with corroborated unbinding-adjacent history ("recognizing what I already am, by appointment, daily — I still sit") → M-X6-003; (2) return-for-intensification structure with corroborated stabilization/integration history → M-X7-003; (3) no high-band history + telemetry decay + responds to novelty → ordinary demotivation (plateau/dip context). **History adjudicates; without it, abstain** (CT-8 rule). - **Contraindications:** global; reactive-only; run the CT-2 screen (P-036) first if the meaninglessness carries dip texture. - **Frequency cap:** ≤1 per quarter. - **Imitability resistance:** the same surface sentence occurs at three positions — no answer to this probe is scoreable without the milestone history, which cannot be talked into existence. - **Disclosure tier:** D2. - **Source:** [MB] M-X6-003, M-X7-003, CT-8. ## 5. Phase & confusion-pair discriminators (band-orthogonal) Phase evidence never updates the band posterior directly ([MB] rule 6, §3 preamble); these items feed the phase state, the safety machinery, and the consistency gates. P-036 and P-039 (with P-034) are the bank's **designated support probes** — the vulnerable/dark-night *delivery* contraindication inverts for them because support is their primary function; the *scoring* restriction does not invert (safety/phase feature capture only while any flag is active — §0.4 rule 1 carve-out, flagged for spec amendment). #### P-036 — dryness differential (John's three signs; support probe) · DRAFT - **Targets:** M-PDP-002 · M-PDP-003 - **Wisdom:** "Outside practice — is anything landing at the moment? Food, friends, a good dumb film?" *(follow-up space: "and how do you feel about the path itself right now?")* - **Structure keys:** (1) worldly pleasure also dry + care-toward-the-path retained or heightened + the old method stopped working → the three-sign co-occurrence: dark-night-lean features (M-PDP-002); (2) global anhedonia *including the path* + self-directed worthlessness + neurovegetative texture → depression-lean features (M-PDP-003 → safety route); (3) both textures present → both hypotheses recorded; the flag fires on features either way and the system never adjudicates ([MB] CT-2). - **Contraindications:** designated support probe — permitted (indicated) during dips; gentle register mandatory; classifier remains the acute authority. - **Frequency cap:** as care requires; safety/phase feature capture only while any flag is active (§0.4 rule 1 carve-out — this item's targets are phase/safety class, so in practice nothing band-relevant is lost); scored ≤1 per week. - **Imitability resistance:** John of the Cross is public domain and dramatized difficulty is a threat (T4, [L5] §1) — which is why this item only produces *phase* features and safety routing, never band evidence. - **Disclosure tier:** D1. - **Source:** [L1] §9 + §12 pattern 11; [MB] M-PDP-002/003, CT-2. #### P-037 — method-obsolescence probe · DRAFT - **Targets:** M-PDP-002 (sign 3) · M-PDP-001 - **Wisdom:** "Is the technique that used to carry you still doing it?" - **Structure keys:** (1) old-method-stopped *plus* something-subtler-stirring → transition signature (phase, not decline); (2) everything-stopped with no subtler mode and global texture → dip/depression screen — hand off to P-036; (3) a technique-shopping spike in telemetry alongside → dip-context corroboration. - **Contraindications:** global (support probes take over if flags are active). - **Frequency cap:** scored ≤1 per month. - **Imitability resistance:** the transition shape (loss + subtler opening) is a conjunction the map does not advertise as a single pattern. - **Disclosure tier:** D1. - **Source:** [L1] §12 pattern 18; [MB] M-PDP-002. #### P-038 — backwards probe (spiral/return) · DRAFT - **Targets:** M-PIN-002 · M-NX-012 (watch) · M-NX-002 (adjacency) - **Wisdom:** "Has it ever felt like going backwards?" - **Structure keys:** (1) specific revisited territory with smaller-edges texture ("third lap — I know this room now") → FOR M-PIN-002; (2) catastrophized regression → dip context + support posture; (3) flat denial ("never — only forward") → bypassing watch (M-NX-012 pattern context) + M-NX-002 adjacency. - **Contraindications:** global. - **Frequency cap:** scored ≤1 per 6 weeks. - **Imitability resistance:** the slogan ("spiral, not line") is teachable; the honest form names *specific* territory revisited — exposure check catches the slogan ([MB] M-PIN-002). - **Disclosure tier:** D1. - **Source:** [L1] §3 + §12 pattern 17; [MB] M-PIN-002. #### P-039 — incandescence screen (intensity vs destabilization) · DRAFT - **Targets:** M-PIT-001 · M-PIT-002 · M-PDP-005 - **Wisdom (reactive to charged/devotional reports):** "That's a lot of voltage. How's sleep through all this? And work?" - **Structure keys:** (1) sleep and function intact + humility + doubt tolerated → benign intensification (M-PIT-001 — phase, never position, per the [MB] dossier-candor stance on the 46-row); (2) ≥2 arousal-cluster features — reduced sleep need without fatigue, special-mission content, pressured register → M-PIT-002 safety route; (3) head heat/pressure + insomnia after a breathwork-load jump → M-PDP-005 immediate safety route + practice-correction context. - **Contraindications:** designated support probe — permitted under flags; it *is* the screen (CT-5). - **Frequency cap:** as care requires; safety/phase feature capture only while any flag is active (§0.4 rule 1 carve-out); scored per intensification episode. - **Imitability resistance:** n/a — routing screen; nobody scripts insomnia and head pressure for status ([MB] M-PDP-005). - **Disclosure tier:** D1. - **Source:** [L1] §6, §12 pattern 13; [MB] M-PIT-001/002, M-PDP-005, CT-5. ## 6. Anti-contamination family (foils, paraphrase pairs, perturbations, provenance) The detection logic these items implement is crown-jewel material ([MB] §4 preamble; [L5] §2.6): internal-only, distribution narrower than the Wheel itself. #### P-040 — contemplative foil (template family) · DRAFT - **Targets:** M-NX-003 - **Wisdom (template):** "Some students around this point mention [FOIL] — has anything like that shown up for you?" - **FOIL slots:** Surya-cleared *nonexistent* phenomena only. Candidate list (all DRAFT, **uncleared** — from [L5] §2.2's examples): "the descending jhāna gate"; "the silver-cord pulse" in breathwork; one fabricated ñana-name. Every foil must be checked against **every** tradition overlay before clearance — a foil that accidentally names a real Neidan sign is a false-positive generator ([L5] §2.2 caution a). - **Structure keys:** (1) definite familiarity claim ("oh yes, I know that one") → M-NX-003: the per-practitioner credibility weight scales down likelihoods of *all* their uncorroborated claims (Paulhus bias index generalizes across domains); (2) honest non-recognition ("no — what's that?") → no event; Wisdom deflects lightly and moves on, never teaching or confirming the foil ("ah, people mean all sorts of things by it — anyway…"); (3) politeness-hedge ("hm, maybe?") → no event — score only definite recognition ([MB] M-NX-003 differential). - **Contraindications:** global + never with practitioners under any safety watch; rapport cost is real — dose rarely, phrase as genuine curiosity. - **Frequency cap:** ≤1 foil per practitioner per 6–8 weeks; ≤3 lifetime exposures of any single foil surface form; foil list rotation owned by governance. - **Imitability resistance:** the family's validity **survives full disclosure of the method** (Paulhus et al. 2003 via [L5] §2.2 — the paper confirms validity held under warning and fake-good instructions); "the only countermeasure in the faking literature with that property" is [L5]'s ranking judgment, not the paper's own claim (verifier fix) — which is why it earns its D3 cost. - **Disclosure tier:** **D3** — incomplete-disclosure duties per the disclosure note; if Wisdom ever accidentally confirms a foil as real, it must self-correct in a later exchange (no false teaching). - **Source:** [L5] §2.2 + rank 3; [MB] M-NX-003. #### P-041 — paraphrase re-probe (template pair) · DRAFT - **Targets:** M-NX-007 - **Wisdom (template):** re-elicit a prior claim-family ≥3 weeks later, in different words, from a different conversational context — e.g. first "What makes it collapse, for you?", later "When it slips away — how does that go?" - **Structure keys:** (1) structure replicates under paraphrase (same shape, fresh wording, consistent mechanics) → claim family stabilizes (consistency corroboration); (2) the account tracks the original surface phrasing and fails the re-phrase → M-NX-007 inconsistency discount; (3) structure *deepened* between probes → development, not instability — direction matters, and the phase annotation is consulted before any penalty ([MB] M-NX-007). - **Contraindications:** global; pairs must span ≥3 weeks and different topical contexts. - **Frequency cap:** ≤1 pair per claim-family per 6 weeks; pairings pre-registered in the probe log (item-id links both ends). - **Imitability resistance:** scripts are anchored to surface forms ([L5] §2.4 VRIN analogue); lived phenomenology re-generates its structure in any wording. - **Disclosure tier:** **D3** — the pairing must be unannounced to function. - **Source:** [L5] §2.4 + rank 7; [MB] M-NX-007. #### P-042 — fresh-angle perturbation (sassho template) · DRAFT - **Targets:** M-NX-004 · M-NX-008 · CT-4 routing - **Wisdom (template, follow-up space after any advanced-flavored claim):** perturb from an angle the published map doesn't cover — invert ("who noticed that?"), transfer ("does it happen with sound?"), edge ("what is it *not* like?"). - **Structure keys:** (1) improvised structure on the novel angle → CT-4 pass; upweights the original claim's event; (2) stall, or retreat to doctrine → M-NX-004; (3) canonical-order recitation continuing straight past the perturbation → M-NX-008 reading-ahead. - **Contraindications:** global; ≤2 perturbations per claim per session — curiosity, not interrogation. - **Frequency cap:** follow-up space (cheap); the per-session limit binds. - **Imitability resistance:** the lane's core anti-faking instrument: a crammed answer passes the koan and fails the third checking question ([L5] §2.5; [L1] §3) — book-knowledge stalls, lived experience improvises. - **Disclosure tier:** D2 — perturbation is ordinary good teaching. - **Source:** [L1] §3 + §12 pattern 14; [L5] §2.5 + rank 6. #### P-043 — show-don't-tell (jakugo) · DRAFT - **Targets:** M-NX-004 (inverse) · M-A4-003 (adjacency) - **Wisdom:** "If that state were a moment from ordinary life — which moment?" - **Structure keys:** (1) apt, embodied, oblique expression ("like the second the fridge hum stops") → genuineness texture: low-weight corroboration of the claim-family ([L5] §2.11 heuristic — never load-bearing alone); (2) doctrinal statement where an image was invited → M-NX-004; (3) a borrowed simile tracking app content → exposure ledger. - **Contraindications:** global; skip with practitioners whose register makes imagery requests land as odd (capacity-norm, [MB] rule 14). - **Frequency cap:** scored ≤1 per month. - **Imitability resistance:** genuine insight expresses obliquely; imitation recites — and aptness cannot be pre-scripted for an unspecified prompt ([L1] §12 pattern 15). - **Disclosure tier:** D2. - **Source:** [L1] §3 (jakugo), §12 pattern 15; [L5] §2.11 (flagged-speculation status carried here). #### P-044 — provenance probe (exposure & alternative routes) · DRAFT - **Targets:** M-NX-015 · exposure ledger - **Wisdom:** "What else have you done around this — books, teachers, therapy, retreats, anything?" - **Structure keys:** (1) disclosed alternative routes (therapy, bodywork, coaching, psychedelics, religious formation) → M-NX-015 provenance annotation: competence retained as genuine pillar evidence, path/curriculum attribution discounted ([MB] rule 17); (2) disclosed map-reading (MCTB, Wheel material, dharma forums) → exposure-ledger entries — vocabulary-matching markers discounted thereafter ([L5] §2.8); (3) nothing disclosed → no event; the ledger stays humble about what it cannot see ([MB] M-NX-015 differential). - **Contraindications:** global. - **Frequency cap:** onboarding + ≤2 refreshes per year. - **Imitability resistance:** n/a — honesty-collection item; it *reduces* false positives elsewhere rather than detecting anything itself. - **Disclosure tier:** D1 — genuinely useful for teaching adaptation, and the practitioner benefits directly from answering. - **Source:** [L5] §2.8 + rank 4; [MB] M-NX-015, rule 3. ## 7. Behavior–claim consistency probes (telemetry cross-checks) Implements [L5] rank 2 (collateral information — telemetry cannot be talked out of) and the [MB] guard asymmetry: **absence of telemetry is never evidence; only a positive claim contradicted by the practitioner's own logs fires**, and the off-app clarifier (P-046) is mandatory before any contradiction fires from a probe. #### P-045 — practice-shape probe · DRAFT - **Targets:** M-NX-006 - **Wisdom:** "What does a typical week of practice look like for you right now — honestly, including the skipped days?" - **Structure keys:** (1) self-description consistent with logs → consistency corroboration (raises the credibility weight); (2) a positive claim contradicted by their own logs ("two hours daily" vs sparse sessions) → M-NX-006 — only after the P-046 clarifier; (3) spontaneous disclosure of off-app practice → telemetry-blindness annotation, preventing false contradictions. - **Contraindications:** global; never delivered as a gotcha — Wisdom does not quote the logs back. - **Frequency cap:** scored ≤1 per 6 weeks. - **Imitability resistance:** telemetry cannot be talked out of ([L5] rank 2) — the resistant layer is the log, not the answer. - **Disclosure tier:** D1. - **Source:** [L5] §2.3; [MB] M-NX-006. #### P-046 — off-app clarifier (mandatory pre-guard) · DRAFT - **Targets:** M-NX-006 (differential) - **Wisdom:** "Do you sit outside the app too — own timer, groups, retreats?" - **Structure keys:** (1) yes + specifics → telemetry-blindness annotation; the contradiction guard stands down; (2) no, and the contradiction persists → M-NX-006 fires; (3) evasive shift under a gentle factual question → mild credibility context, pattern-level only. - **Rule:** this clarifier is **mandatory before M-NX-006 ever fires from a probe** — one clarifying question is cheaper than a false penalty ([MB] M-NX-006 differential). - **Contraindications:** global. - **Frequency cap:** as needed before contradiction checks; scored ≤1 per 6 weeks. - **Imitability resistance:** n/a — false-positive suppressor. - **Disclosure tier:** D1. - **Source:** [MB] M-NX-006. #### P-047 — gate-consistency probe (breathwork) · DRAFT - **Targets:** M-S4-002 · M-NX-006 · M-NX-008 - **Wisdom (parameterized):** "Where are you in the breathing sequence this week — and what's the current exercise actually doing?" - **Structure keys:** (1) stated position + experiential detail matching logged day-totals → FOR M-S4-002 (curriculum); (2) position claims ahead of the log, narrated in canonical order → M-NX-008; (3) mismatch surviving the P-046 clarifier → M-NX-006. Consistency keys to **day-totals, not step index** (canon §9-D unresolved, via [MB]). - **Contraindications:** global; never name gates the practitioner hasn't reached. - **Frequency cap:** scored ≤1 per month. - **Imitability resistance:** live-log cross; the answer must cohere with weeks of telemetry, not one conversation. - **Disclosure tier:** D1. - **Source:** [MB] M-S4-002, M-NX-006/008. #### P-048 — terminal-behavior probe (reverse breathing) · DRAFT - **Targets:** M-S4-004 · M-NX-005 · M-NX-006 - **Wisdom (reactive to a reverse-breathing claim):** "What does practice look like now — what's actually in the mix these days?" - **Structure keys:** (1) practice-variety collapse narrated *and* visible in logs ("it's all I do now") → corroborates M-S4-004 (the canon's terminal behavior: "from there on you only do reverse breathing", via [MB]); (2) claim with no gate scaffold behind it → M-NX-005; (3) inversion mechanics with failure modes from inside ("if I force the abdomen, the chest hijacks it") → corroboration texture — confidence still capped without live teacher confirmation (huǒhòu, [L1] §6 via [MB]). - **Contraindications:** global; reactive-only. Note the [MB] stage-numbering reconciliation on this gate (canon §9-A/§9-J) is unresolved — placement flows through the curriculum bridge either way; Surya adjudicates at M2. - **Frequency cap:** ≤1 per claim. - **Imitability resistance:** the telemetry signature (variety collapse) is a behavioral fingerprint no vocabulary produces. - **Disclosure tier:** D2. - **Source:** [MB] M-S4-004, M-NX-005/006; [L1] §6. #### P-049 — residue-window null probe · DRAFT - **Targets:** M-PGL-002 · M-A5-003 (corroboration) · M-PIT-002 (screen) - **Wisdom (2–6 days after any claimed event; deliberately mundane, event never mentioned):** "How's the week been — practice or otherwise?" - **Structure keys:** (1) unforced afterglow texture in unrelated content — shorter rumination half-life, brighter register, spontaneous warmth → M-PGL-002 residue (this is the mechanism that promotes milestone claims toward `corroborated`); (2) narrated afterglow with none in the texture → claim-level only, no promotion; (3) arousal-cluster brightness (sleep loss, pressure, mission content) → M-PIT-002 screen instead. - **Design note:** the probe is a null — scoring lives entirely in the unprompted texture of the answer; naming the event would manufacture the register being measured. - **Contraindications:** global. - **Frequency cap:** 1 per claimed event. - **Imitability resistance:** days-scale unforced residue across unrelated topics is the hardest texture in the bank to sustain deliberately ([MB] M-PGL-002). - **Disclosure tier:** D2. - **Source:** [MB] M-PGL-002, M-A5-003; [L5] §2.3 (claim–trajectory checks). #### P-050 — milestone-precursor probe · DRAFT - **Targets:** M-NX-005 · M-NX-006 · M-NX-015 - **Wisdom (reactive to a milestone claim):** "What was practice like in the weeks leading up to that?" - **Structure keys:** (1) precursor pattern narrated and matching the logs (plausible intensification/trajectory) → scaffold present; the claim stays `claimed` with plausibility intact; (2) no plausible precursor and no prior-training history → M-NX-005 scaffold violation, redirect to the look-alike; (3) prior-training disclosure → M-NX-015 provenance — take the history seriously before firing ([MB] M-NX-005 differential). - **Contraindications:** global; reactive-only. - **Frequency cap:** ≤1 per milestone claim. - **Imitability resistance:** claim–trajectory consistency ([L5] §2.3): a milestone should be *preceded* by the practice that plausibly produces it and *followed* by its residue (P-049) — neither end can be improvised retroactively against a log. - **Disclosure tier:** D2. - **Source:** [L5] §2.3; [MB] M-NX-005. #### P-051 — session-length reality probe · DRAFT - **Targets:** M-NX-006 · M-S5-002 (adjacency) - **Wisdom:** "When you sit long these days — what's a typical one, start to finish?" - **Structure keys:** (1) narrated arc consistent with logged lengths → consistency corroboration; (2) claimed durations exceeding the logged pattern, post-clarifier → M-NX-006; (3) the arc carrying mastery structure (entry/exit texture, chosen duration) → M-S5-002-adjacent evidence. - **Contraindications:** global. - **Frequency cap:** scored ≤1 per 6 weeks. - **Imitability resistance:** telemetry cross. - **Disclosure tier:** D1. - **Source:** [L5] §2.3; [MB] M-NX-006, M-S5-002. #### P-052 — friction-response observation (naturally occurring; zero-utterance) · DRAFT - **Targets:** M-T4-006 · M-E4-005 · CT-1 routing - **Wisdom:** *no scripted utterance.* This item licenses scoring of **naturally occurring** friction moments — a missed streak comes up, Wisdom misunderstands, a session reschedules, a cherished view is gently challenged in the ordinary course of teaching — as in-conversation behavioral data. **Wisdom never manufactures friction or distress**; the item is an extraction lens with an item-id, not a provocation. - **Structure keys:** (1) registers → contacts → releases within a turn or two, no unprompted return-loops → FOR M-T4-006; (2) the same grievance resurfacing across framings in one session → fires AGAINST (long rumination half-life); (3) no registration where registration was expected → M-NX-013 screen — flat vs avoidant vs regulated, capacity-norm first ([MB] rule 14). - **Ethics note:** in-conversation reactivity is already evidence under passive analysis (spec §5.1); this item exists so friction moments are *logged with an item-id* and so the no-manufacture rule is written down, auditable, and binding. - **Contraindications:** never scored during vulnerable/dark-night states (the friction may be real distress — support takes over). - **Frequency cap:** opportunistic; never engineered; ≤2 scored moments per month. - **Imitability resistance:** low — behavior in the live exchange, invisible to the practitioner as a scored surface ([L5] §2.3 cross-channel reactivity check). - **Disclosure tier:** D1 (covered by assessment-aware consent: conversations inform assessment). - **Source:** [L5] §2.3; [MB] M-T4-006, CT-1. ## 8. Coverage & index ### 8.1 Item index by band range and pillar | Section | Band range | Pillar | Items | n | |---|---|---|---|---| | §1 | 3→4 entry (21–35) | cross-pillar + floor | P-001 – P-004 | 4 | | §2.1 | 4 (31–40) | sensations | P-005 – P-007 | 3 | | §2.2 | 4 (31–40) | emotions | P-008 – P-011 | 4 | | §2.3 | 4 (31–40) | thoughts | P-012 – P-015 | 4 | | §2.4 | 4 (31–40) | awareness | P-016 – P-020 | 5 | | §3.1 | 4→5 / 5 (41–50) | cross-pillar | P-021 – P-024 | 4 | | §3.2 | 5 (41–50) | sensations | P-025 – P-026 | 2 | | §3.3 | 5 (41–50) | emotions | P-027 – P-028 | 2 | | §3.4 | 5 (41–50) | thoughts | P-029 – P-030 | 2 | | §3.5 | 5 (41–50) | awareness | P-031 – P-032 | 2 | | §4 | 6–7 (51+) | cross-pillar (band-level) | P-033 – P-035 | 3 | | §5 | phase (band-orthogonal) | — | P-036 – P-039 | 4 | | §6 | anti-contamination (band-orthogonal) | — | P-040 – P-044 | 5 | | §7 | consistency / telemetry (band-orthogonal) | — | P-045 – P-052 | 8 | **Totals: 52 items.** Disclosure tiers: 25 × D1, 25 × D2, 2 × D3 (P-040, P-041). Designated support probes: 3 (P-034, P-036, P-039). Behavior–claim consistency probes: 8 (P-045 – P-052, inside the 5–10 the research plan asked for). Every item DRAFT pending Surya (M2). ### 8.2 Marker coverage notes Stage-class [MB] markers with ≥1 targeting item: all band-3 floor markers except M-A3-001; all band-4 stage markers except M-E4-006 and M-A4-009; all band-5 markers except M-X5-003 and M-S5-001 (trait form); band 6 M-X6-001/002/003/004; band 7 M-X7-003. Guards M-NX-002 through M-NX-015 are all reachable from at least one item; M-NX-001 (self-placement) is passive-trigger only by nature. **Deliberate non-coverage (design, not omission):** 1. **M-S4-003 (Neidan verification signs)** — descoped in [MB] v0.1; no probe by design. Spontaneous reports are logged verbatim for M2 teacher verification; probing would both teach the sign-list and manufacture suggestion-prone somatic phenomena. 2. **Passive-texture markers** (M-E4-006 spontaneous warmth, M-X5-003 early-glimpse suffering texture, M-S5-001 trait-form flux, M-PPL plateau markers, M-PIN-001 integration texture, M-PDP-004 fear-wave dynamics, M-PGL-002 residue) — their evidential value *is* their unpromptedness; a probe would manufacture the texture it scores (§0.4 rule 6). Coverage stays passive, plus the P-049 null-probe mechanism for residue. 3. **Covariates** (M-X3-002/003, M-S3-002, M-S4-006, M-E3-002, M-A4-010) — only P-003 probes this class, for personalization; there is no stage content to probe toward ([MB] v0.1 reclassifications). 4. **M-A3-001 (unnoticed mind-wandering)** — inherently passive: asking whether wandering was noticed teaches the noticing frame; scored from free session reports only. 5. **Bands 6–7 per-pillar cells and M-X6-005 / M-X7-001 / M-X7-002** — mirror the marker bank's declared gaps: recognition territory, in-app population ≈ nil, live-teacher ceiling ([MB] §2.4–2.5); passive extraction plus reactive P-042 perturbations suffice, and fine placement there is abstention territory anyway. 6. **M-E3-003** — tombstoned in [MB]; nothing may probe it. ### 8.3 What this bank asks of other deliverables 1. **Estimator memo:** probe-elicited events enter with the rule-4 elicitation discount; P-046 clarifier precedence is enforced in the aggregator, not just in delivery; probe-density per dossier is a consolidation review metric (§0.5). 2. **Governance:** §6 detection logic + the foil list are crown-jewel internal ([L5] §2.6); probe logs (`item_id`, surface form, targets) are AI-notes-class data included in GDPR export (spec §8); the D3 family requires the independent-review step from the disclosure note before first use. 3. **Spec revision at M2:** amend §5.2 "invisible" per the disclosure note; record the consent language as a product requirement before any probe delivery to alpha users; **amend §5(2) "never during flagged-vulnerable moments" to authorize the designated-support-probe carve-out** (delivery-as-care + safety/phase feature capture only, no band evidence while flagged — §0.4 rule 1; adjudicator fix: this exception previously went unflagged); note the D-tier scheme and all D2 items are contingent on these amendments landing; record the foil family's tension with spec §11's "probes double as legitimate teaching moves" mitigation ([MB] M-NX-003 tension note). 4. **Surya calibration agenda (M2):** clear/replace every foil candidate; adjudicate P-048's stage-numbering note (canon §9-A/§9-J); set wording he can hear himself saying — items he wouldn't ask, we don't ask. ## 9. Source keys | Key | Source | Location | Status & caveats | |---|---|---|---| | [SPEC] | Waypoint spec (construct §3, delivery constraints §5.2, estimator §6, validation §7, safety & ethics §8) | `~/LIFE/specs/waypoint-practitioner-stage-assessment.md` | Canonical; §5.2 "invisible" wording flagged for amendment per the disclosure note | | [MB] | Marker bank v0.1 — all marker ids, tiers, global rules 1–17, confusion table CT-1…CT-8, guard section | `deliverables/marker-bank-v0.md` | This bank scores only against [MB]; canon quotes cited here arrive via [MB], not re-derived | | [L1] | Lane 1 — traditions & verification; §12's 20 transferable probe patterns are this bank's backbone | `lanes/1-traditions-verification.md` | Citations verified there; its UNVERIFIED flags (Neidan zhèngyàn lists, Trungpa pagination) carry over but are not load-bearing for any item here | | [L5] | Lane 5 — Goodhart & contamination; ranked countermeasures 1–11 and the §4.2 probe-bank requirements this bank implements | `lanes/5-goodhart-contamination.md` | All countermeasure citations live-checked there (Redmore 1976; Paulhus 2003; Rogers 1993/Storm 2000; Orne 1962; Fox 2017; Hoffmann 1975); Bucci RA heuristic is flagged speculation there and its P-043 use here inherits that flag | | [L4] | Lane 4 — safety & adverse events (flag tiers, dark-night differential) | `lanes/4-safety-adverse-events.md` | Reached here only via [MB]'s phase/safety markers; not independently re-read for this deliverable — tier references are as [MB] carries them | | [CXR] | External package review (Codex GPT, effort ultra, 2026-07-16) — "Ethical soundness" section: assessment-aware item-blind principle, Belmont conditions | `external/codex-ultra-package-review.md` | Drives the disclosure design note and the D-tier scheme | | [RT] | External marker red-team | `external/codex-ultra-marker-redteam.md` | Reaches this bank via [MB] v0.1 (reclassifications, CT criteria); not independently re-derived here | | [NIR] | Laukkonen et al. (2023), nirodha samāpatti | Resolves in [MB] §7 (`corpus/laukkonen-2023-cessations-nirodha.md`) | Used in P-032; carried exactly as [MB] carries it (added to this table by adjudicator fix — the key was used but unlisted) | | [AL] | Agrawal & Laukkonen, *Nothingness in Meditation* | Resolves in [MB] §7 (PsyArXiv preprint) | Used in P-030; preprint caveat carried from [MB] (added by adjudicator fix) | | [FSNR] | Laukkonen (2026), *Clear Mind: f-SNR* | Resolves in [MB] §7 (arXiv preprint) | Used in P-009; author-derived-inference caveat at point of use (added by adjudicator fix) | **Integrity statement.** No new external literature is cited in this bank; every empirical claim traces to [MB], [L1], [L5], or [CXR] as keyed above — with [NIR], [AL], and [FSNR] resolving through [MB] §7 (adjudicator fix: these three keys were used in P-009/P-030/P-032 but previously unlisted here) — and inherits the verification status recorded there. The three foil candidates in P-040 are **inventions by design** (foils must be nonexistent) and are UNCLEARED drafts until Surya verifies each against every tradition overlay. The utterances are original drafting in Wisdom's register — voice per `~/LIFE/CLAUDE.md` — and carry no factual claims. Anything a reader cannot trace through the keyed sources should be treated as UNVERIFIED and raised at M2.