04 · The use-case catalogue

A feature is a pipeline configured for a job (01 §6). This document enumerates the jobs. Each entry names who has the job, what the theory lets us compute for it, what goes in and out, how we would know it works, its build status, and who pays. The catalogue is deliberately over-complete — prioritisation is the roadmap's task, not the catalogue's.

How to read an entry

Every use case is specified on eight fields: User (who), Job (the outcome they want), Hook (the corpus result that makes it computable), In/Out (evidence in, readout out), Validation (how we confirm it works), Status ([design] buildable now · [research] needs a bet · [speculative] frontier), Pay (the monetisable surface). Anchor classes (I/II/III) are named where they matter.

§1. Individual — self-reading and navigation

1.1 State read. User: anyone doing self-work. Job: "where am I, really — as coordinates, not a horoscope." Hook: the eleven invariants (02). In: Anchor-II self-audit (or I where a wearable is present). Out: mandalagram + invariant readout + archetype + mode. Validation: convergent validity against Big Five, well-being, and integration/differentiation scales; test-retest reliability. Status: [design]. Pay: consumer subscription (the core loop).

1.2 Trajectory & personal forecast. User: returning user. Job: "what's coming — when will this channel open, when am I near a break." Hook: transparency windows, turbulence zones, bifurcation warnings. In: autoephemeris (≥4 sessions). Out: trend chart, window/turbulence forecast, saddle-node ("dark night") early warning. Validation: prospective — does a flagged bifurcation precede a self-reported crisis above chance? Status: [research] (needs longitudinal data). Pay: premium tier; the retention engine.

1.3 Correction / practice guidance. User: someone stuck on a specific channel. Job: "smallest thing I can do to reopen X." Hook: minimal-intervention protocol — which the corpus reads as ethics itself: self-correction of the Gap-profile toward alignment, since minimising total Gap is equivalent to maximising stable purity (value-consciousness §optimality of alignment [C]). In: current Gap profile. Out: the minimal-intervention target + optimal frequency. Validation: RCT-style — does the recommended intervention move the target channel faster than a control practice? Status: [research]. Pay: coaching add-on; partner practitioners.

1.4 Decision support. User: someone facing a choice. Job: "which option keeps me alive to more of myself." Hook: Meaning vector + Freedom (flat directions), constrained by the value hierarchy — vital $\succ$ homeostatic $\succ$ social $\succ$ cognitive $\succ$ aesthetic $\succ$ transcendent, so an option that buys a higher-tier gain by spending viability is never surfaced as an improvement. In: two hypothetical Γ-shifts. Out: projected change in Meaning and in Freedom per option, with any value-hierarchy violation flagged. Status: [design] (as a structured reflection tool, labelled II). Pay: premium.

1.5 Alignment — the knowing–doing gap. User: anyone whose actions diverge from their knowledge or values. Job: "where am I at war with myself, and what does it cost." Hook: misalignment is a computable quantity — $\mathrm{Gap}(L,D) = |\sin(\arg\gamma_{LD})|$ (knowledge vs action), with the corpus result that high misalignment lowers purity (value-consciousness §misalignment [C]). In: Anchor-II/I estimate. Out: the misalignment map across L–D, E–U, and the value channels, with the highest-cost discord named. Status: [design]. Pay: premium; the honest, measured version of "inner-conflict" coaching.

§2. Dyadic — synastry, done honestly

2.1 Compatibility / synastry. User: couples, co-founders, collaborators. Job: "how do we actually fit." Hook: composite Γ; the cooperation theorem [T] — shared cross-coherence strictly raises joint purity, $P(\rho^{(12)}_*) = P(\rho_{\mathrm{diag}}) + 2\lVert\gamma_{\mathrm{cross}}\rVert_F^2$ (value-consciousness); empathy $= 1 - \mathrm{Gap}_{AB}(E,E)$ and the golden rule as the symmetry of $\Gamma_{\mathrm{composite}}$ (value-consciousness); the one-theme law. In: two Anchor-II estimates. Out: cross-coherence map (which channels connect), the empathy score (E-sector coherence), the cooperation gain $2\lVert\gamma_{\mathrm{cross}}\rVert_F^2$ (the measurable "free purity" of working together), and the jam detector — two circuits sharing two themes is a theorem-grade locked conflict, not a vibe. Validation: predictive of relationship-satisfaction measures; the jam pattern should correlate with reported recurring conflict; the cooperation gain should track joint-task performance. Status: [design]. Pay: the highest-converting consumer surface (relationship products always are), and B2B for co-founder/hiring fit.

2.2 Mediation. User: a therapist/mediator. Job: "name the channel the conflict flows through." Hook: the bridge table (one-theme law gives the unique shared channel of any two circuits). Out: the named bridge + its transparency. Status: [research]. Pay: practitioner licence.

§3. Collective — teams and organisations

3.1 Team profile. User: team lead. Job: "what is our collective state and where is it weak." Hook: conciliar mandalagram + 7-D org profile. In: team members' σ-audits. Out: org-scale $P,\Phi$, Gap map, the weakest circuit. Validation: against team-performance and engagement metrics. Status: [design]. Pay: B2B seats — the highest-margin surface.

3.2 Org-pathology detection. User: org consultant. Job: "find the structural conflict before it blows up." Hook: the one-theme law as a pathology theorem — two departments sharing two functions is a jammed jurisdiction. Out: the jam list, ranked. Status: [design]. Pay: consulting toolkit licence.

3.3 Team formation. User: org designer. Job: "compose a team whose circuits share exactly one function each." Hook: the same law, run forward as a constraint. Status: [research]. Pay: enterprise.

§4. Clinical and health — hardest, most valuable

4.1 Consciousness-level assessment. User: anaesthesiologist / ICU / DOC clinician. Job: "is there someone in there, and how deep." Hook: altered-states profiles on the PCI bridge ($P_{\mathrm{crit}}=2/7 \leftrightarrow \mathrm{PCI}=0.31$). In: Anchor-I neural signals. Out: $P$ with confidence, band membership. Validation: against established DOC/anaesthesia-depth measures — the corpus prediction F-Neural is exactly this. Status: [research] → regulated [medical]; the highest-stakes and highest-validation surface. Pay: clinical device/software (regulated).

4.2 Meditation-depth staging. User: contemplative / researcher. Job: "objectively stage this practice." Hook: the samādhi/shamatha profiles + the training law (baseline $R_\varphi$ rises with practice — a measurable curve). Status: [research]. Pay: research licence; premium contemplative tier.

4.3 Mental-health trajectory (adjunct, not diagnosis). User: therapist + client. Job: "track state between sessions." Hook: autoephemeris + bifurcation warning. Status: [research], explicitly non-diagnostic until validated. Pay: practitioner tier.

§5. Contemplative — practice instrumentation

5.1 The practice log. User: any practitioner. Job: "see whether my practice is actually changing my baseline." Hook: the two-timescale training law — a falsifiable progress curve, not a subjective claim. Status: [design] (log) → [research] (curve validation). Pay: premium.

5.2 The oracle, honest. User: someone who uses casting. Job: "a structured lens on a question." Hook: the П4 protocol, labelled Anchor III. Status: [design]. Pay: free/engagement feature — the honest replacement for the tarot draw.

§6. AI and alignment — the same predicate on silicon

6.1 Agent introspection & alignment. User: AI lab / safety team. Job: "measure an agent's integration, reflection, viability — and its alignment." Hook: substrate closure; the SYNARC in-silico confirmation is the proof of concept. Crucially, "alignment" is not a metaphor here: the corpus's misalignment is exactly $\mathrm{Gap}(L,D)$ — the phase discord between an agent's knowledge (L) and its action (D) — a computable quantity that provably lowers purity (value-consciousness §misalignment [C]). In: agent telemetry → Anchor I. Out: $P,R,\Phi$ on the agent over training/inference, plus the misalignment map (does the agent act on what it knows, and where does it not). Validation: against known failure episodes (the hallucinating-LLM case); rising $\mathrm{Gap}(L,D)$ should precede knowing-but-doing-otherwise failures. Status: [research], but immediately fundable. Pay: B2B/enterprise — a distinct, well-capitalised market on the same kernel; the alignment readout is the strongest wedge into AI-safety budgets.

6.2 Multi-agent ecology. User: multi-agent system operator. Job: "read the collective state of a swarm." Hook: composite Γ over agents. Status: [research]. Pay: enterprise.

§7. Scientific — the validation platform

7.1 Prediction testing. User: researchers (and the project itself). Job: "test UHM's falsifiable predictions at scale." Hook: the whole instrument. Out: population data on F-Gap-2 (triad clustering), F-ISF, F-Neural. Status: [design] (data collection) — this is where product and science are the same activity (00 §2). Pay: grants; the legitimacy flywheel.

§8. Education — literacy in the frame

8.1 Self-literacy. User: learners. Job: "understand my own qualities in a rigorous frame." Hook: the mandalagram as a teaching object; the bridge/third laws as reasoning exercises. Status: [design]. Pay: education licence.

§9. Symbolic interop — the universal translator

9.1 Back-projection. User: anyone arriving from astrology / I Ching / tarot / Human Design. Job: "show me what my existing reading captured, and what it dropped." Hook: the back-projection protocol + the classification T-256 + the information-loss theorem. In: a legacy reading. Out: its Γ-image, its orbit type, and its quantified loss. Status: [design]. Pay: the onboarding funnel — meet users inside the system they already trust, then reveal the derived object beneath. Strategically the most important acquisition surface.

§10. Prioritisation preview

The roadmap (07) sequences these by (validatability now × market pull × build cost). The natural V0 set is 1.1, 5.2, 9.1 (self-read, honest oracle, back-projection) — all [design], all consumer, all validatable against existing psychometrics. The natural first B2B is 6.1 (AI introspection) — same kernel, well-funded buyer, no consumer-scale UX needed. The highest-stakes long bet is 4.1 (clinical) — where the corpus's sharpest prediction lives and where regulation gates entry.

Where this leads. 05 · The human specialisation develops the individual domain (§1) into the deep, human-specific module the project treats as first-class.