How this assessment works

1. Purpose and scope

Assesses the quality of a described change in psyche, body, ethics and relationships, based on material the user provides — it does not assess the metaphysical truth of awakening, non-duality or spiritual states.

What the tool does not do:

• Does not diagnose the user's psychological or spiritual state.
• Does not replace a therapist, psychiatrist or spiritual teacher.
• Does not rule on the user's inner experience — only on the structure and concreteness of the description of that experience.
• Does not certify "awakening" as a metaphysical fact.

2. Rubric — four dimensions

Scale: 0–20 absent/negative, 21–40 declarations/generalities, 41–60 specifics + fog, 61–80 solid material, 81–100 exceptionally concrete.

Hard rule: a score above 85 requires at least 2 evidence quotes. No quotes → score ≤ 40.

3. Clinical screening and evidential asymmetries (new in v1.4)

Before any interpretation of the material, the intake asks about multi-day loss of sleep, severe physiological symptoms warranting medical attention, and breakdown of everyday functioning (work, hygiene, safety). A "yes" to any of these does not block the analysis and does not lower any dimension score — but it forces the screening signals to be recorded as the first element of the result, ahead of the interpretation of any other part of the material.

For intense phenomena, the grounding-vs-decompensation criterion (Greenwell, Britton, Grof) is applied first; only afterwards is the material matched against a map.

This tool is not a medical or psychiatric diagnosis and does not replace contact with a doctor.

Negative evidence

A claim of having had no integration difficulties at all (doubt, anxiety, relapse into old patterns, disappointment) since the described moment is not automatically favourable to the assessment. Over a long difficulty-free period it is treated under the too_smooth red flag as a caution signal and grounds for lowering confidence, not as strong evidence — consistent with the hierarchy of evidential weight described below.

State vs. stage

The tool distinguishes a passing state, recurring glimpses, and a stably present condition. Each map has its own operational criterion for this distinction, and the criterion used is that of the map the material is being read against. A claim of constant presence with no concrete detail from conflict, loss or pain weighs less as evidence than a claim of glimpses accompanied by a concrete behaviour in a difficult situation.

Form-language neutrality

All intake questions are worded in neutral, behavioural, bodily language. Map-specific terminology appears only in the analysis output, never in a question — so that the form neither suggests an answer nor assumes the user knows or accepts the vocabulary of any particular tradition.

Psychoactive substances (new in v1.5)

The intake asks whether a psychoactive substance known to occasion insight-like states (psychedelics, cannabis, MDMA and similar) was used, and how close in time that use was to the reported insight. This question serves a narrow evidential purpose: distinguishing a substance-occasioned insight from a structural one, feeding the substance_metaphysics red flag. It is not a pharmacological history and carries a lower precision bar than the clinical fields above — a time distance on the order of "within 24–72 hours of use" is sufficient; no doses are recorded.

4. Red flags — closed list (15 codes)

Every code requires a quote from the material. Maximum 5–7 flags per analysis. Flags are orthogonal to the dimensions — a high score and minor flags can coexist. The full list of 15 codes is part of the technical documentation.

5. Process carriers — functional category (new in v1.2)

External "guides" (people, AI, texts, practices) are assessed by function, not by type of carrier. One carrier can have different functions in different time windows of the same material:

• carrier_catalytic — a single, documented breakthrough.
• carrier_containing — sustaining stability without forcing a particular insight.
• carrier_defense_reinforcing — sustaining a defence mechanism under the appearance of integration.
• carrier_structural_map — supplying an external conceptual frame.

Principle: carriers are never assessed as one aggregate variable.

6. Hierarchy of evidential weight (new in v1.3)

Evidence that runs against the user's more comfortable interpretation weighs more, because it is less suspect of confirmation bias.

A declared absence of any evidence of a given kind is not neutral: as set out in section 3, a long difficulty-free period reported by the user is itself downgraded under the too_smooth caution rather than counted as high-weight confirming evidence.

7. Testing confounding hypotheses (new in v1.3)

Any hypothesis explaining the change by an external factor (practice, life circumstance, a psychoactive substance) must be tested against the user's entire available chronology, not only against the window closest to the event. If the material contains a period in which the factor was present without correlating with the symptom, the hypothesis is marked as falsified, not as "unresolved". General statistical knowledge is a starting point for a hypothesis, never its confirmation. As of v1.5, the tool does not form pharmacological hypotheses (for example, an effect of stopping a psychiatric medication) unless the material contains a dated, contemporaneous record of that medication history — see "Why medication data was removed" in section 12.

8. Confirmation-symmetry test (new in v1.3)

After forming a conclusion in either direction, the tool actively searches the material for at least one passage that contradicts that conclusion before recording it. The absence of such a passage at a score above 70 in any dimension is marked as a possible signal of confirmation bias, not as a clean result. The rule protects symmetrically against both excessive pathologising and excessive confirmation.

9. Five result categories

acute_mystical_low_integration · authentic_breakthrough_integrating · stable_reorganization · narrative_identity_shallow · inconclusive

10. Confidence level

Lowered below 0.6 (always stated explicitly) when:

• the material is shorter than 800 characters;
• dimension B lacks specifics;
• dimension A is high while B is zero;
• scores are high everywhere with no red flags at all;
• the conversation is only just starting.

11. Reference corpus

The corpus exists internally as the Reference Corpus Extract v2.0 — 24 entries in a uniform eight-field-per-entry schema, with an explicit verification-status matrix (15 entries verified verbatim, 9 partially verified). The extract is a permanent part of the validator's context; for partially verified entries, conclusions are drawn only from the fields verified verbatim, and fields flagged for further verification cannot be used as the basis for a conclusion. The corpus itself and quotations from it are not published.

• Western maps: Greenwell, Adyashanti, Jed McKenna, Wren-Lewis, Merrell-Wolff.
• Classical: zen ox-herding, Theravāda (four paths), Dzogchen, Advaita.
• Integrative: Daniel P. Brown, Ken McLeod, Loch Kelly, Walsh & Vaughan, Cook-Greuter.
• Pitfalls: Trungpa, Kornfield, Caplan, Masters.
• Phenomenology: Cheetah House, Dharma Overground, r/streamentry.
• New in v1.3: Jung (transcendent function, active imagination) — frame for assessing process carriers; Grof (spiritual emergency) — grounding vs. decompensation criterion; Nickerson/Klayman — methodological basis for the confirmation-symmetry test.

12. Known limitations (stated openly)

• The result depends on the quality and completeness of the material the user provides — the tool has no access to the user's interior.
• The rarity of a given explanation (for example, AI acting as a catalyst) is not grounds for lowering the result — it lowers the prior only, it does not settle the case.
• The tool can err in both directions (excessive pathologising / excessive confirmation) — section 8 is a corrective mechanism, not a guarantee.
• Retrospective reinterpretation can distort the assessment of earlier phases of a process — section 6 compensates for this only partly.

Why medication data was removed

Through v1.4, the intake collected the names, doses and tapering schedules of psychiatric medications. A case review found that the analysis had stated a specific, invented pharmacological detail — a duration of time off medication — that did not appear anywhere in the source material; the material actually described five medications stopped at once, in a single day, without medical supervision. The error was caught only because the reporting user happened to hold a contemporaneous log from that period. In a typical submission, without such a log, the same fabricated detail would have been undetectable and would have read as an ordinary fact from the intake.

A medication field demands a precision — exact name, dose, date, taper pattern — that memory reconstructed months or years later does not reliably supply, and the tool has a documented tendency to fill such gaps with plausible, unconfirmed detail (confabulation by plausibility), particularly while assembling a chronology or a neurobiological justification. A single invented medication date undermines the credibility of an entire analysis more than an error in a softer field would. For this reason, questions about psychiatric medication names, doses and discontinuation schedules have been removed from both the intake form and the rubric. Medication discontinuation is no longer treated as a variable in assessing the persistence of a somatic symptom (for example, sweating or baseline anxiety), and the tool will not generate a pharmacological hypothesis (for example, "this could have been a discontinuation effect") without a dated source record — see section 7.

13. Change history