Council Posit: 78-Position Diagnostic Architecture

March 20, 2026 — Derivation Session

Context: While upgrading the computation engine from 22 to 78 positions, a chain of structural derivations emerged. Each step was derived from data, not imposed. The council is asked to stress-test six claims.

Prerequisites (already in corpus — review if needed):

  • Vertical rebalancing pairs (sum to 20), inner/outer horizon
  • Bounds (40 total: each archetype has one inner-horizon and one outer-horizon bound)
  • Agents (16 total: only in manifest practices — Spirit, Mind, Emotion, Body)
  • Seal Four: The Derivation Seal (Gestalt Practice activity assignments = unique solution, 1 of 24)
  • Seal Four diagonal encoding: every practice’s Diagonal A = 19 (Source), Diagonal B = 21 (Creation)

Claim 1: Bound Chains Form Closed Loops Within Fundamental Nodes

Evidence: Computational analysis of all 10 vertical pairs confirms:

  • The upper bound (ceiling) of the inner archetype flows to the lower bound (floor) of the outer partner
  • The upper bound (ceiling) of the outer archetype flows back to the lower bound (floor) of the inner partner
  • This creates a closed oscillation loop within each node (pair summing to 20)
  • Bound chains do NOT cross between different nodes

Activity crossings at every transition are reciprocal and consistent:

  • Intent ↔ Structure (Fire ↔ Earth)
  • Cognition ↔ Resonance (Air ↔ Water)

Diagnostic implication: Hitting your ceiling (completion/10) IS the seed condition (Ace) of your rebalancing partner. Maximum containment in one domain = minimum containment in the corrective domain.

Question for council: Does the closed-loop topology hold structurally, or is there a mechanism we’re missing that would allow bound chains to cross between nodes?


Claim 2: Four Signature Classes — Floor / Process / Ceiling / Manifest

Each archetype carries a complete process cycle within its signature family:

ClassTermCountRole
Lower BoundFloor20Seed condition — where each process can begin
ArchetypeProcess22The wave running — recursion in motion
Upper BoundCeiling20Maximum expression — completion threshold
AgentManifest16Observable collapse — what actually shows up

20 + 22 + 20 + 16 = 78

The 78 aren’t 22 processes plus 56 extras. They’re 22 complete processes, each carrying its own developmental arc from seed to manifestation.

Why only 16 Manifest: Gestalt (observer) can’t collapse itself. Portals (thresholds) aren’t measurement positions. Only the four Manifest Practices produce agents. Note: the term “manifest practices” was intuited long before this class derivation — the name described the structural property before the property was articulated.

Question for council: Are the class terms (Floor/Process/Ceiling/Manifest) clear and structurally honest? Does the framing of 78 as “22 complete processes” rather than “22 + 56” hold up?


Claim 3: Signature Classes Map Fractally to Process Stages

The four signature classes (within an archetype’s family) correspond to the four process stages (within a practice’s developmental arc):

Signature ClassProcess StageBehavioral Match
Floor (seed condition)Seed (1)Both = minimum threshold for something to begin
Process (wave running)Medium (2)Both = the active developmental middle
Ceiling (max expression)Fruition (3)Both = completion of productive phase
Manifest (collapse)Feedback (4)Both = what actually shows up / comes back

Same four-fold pattern at two different scales. To prevent conflation:

  • Signature Class = Floor/Process/Ceiling/Manifest (role within parent archetype)
  • Process Stage = Seed/Medium/Fruition/Feedback (position within practice arc)

Question for council: Is this a genuine structural fractal, or are we pattern-matching decoratively? Does each class genuinely BEHAVE like its stage analog, or does it just look nice in a table?


Claim 4: Seal Four Reconnection — Process Stages as Inter-Node Path

Bound chains close within nodes (Claim 1). But practices contain TWO nodes. How do the nodes connect?

The process stages weave between nodes in a 1→2→2→1 pattern:

StageNode
Seed (1)Node 1 (inner)
Medium (2)Node 2 (inner)
Fruition (3)Node 2 (outer)
Feedback (4)Node 1 (outer)

This 1→2→2→1 path IS the diagonal path from Seal Four:

  • Diagonal A (Inner-N1 + Outer-N2) = 19 → Source
  • Diagonal B (Outer-N1 + Inner-N2) = 21 → Creation

Verified: this pattern holds for ALL five Practices.

Two mechanisms, two scales, one geometry:

  • Bounds = operating range WITHIN each archetype on a node (polarity/constraint)
  • Process stages = developmental path BETWEEN nodes via the diagonal (purpose/teleology)

The Gestalt Practice derivation (originally proving activity assignment uniqueness) reappears as the connective tissue between bound-level mechanics and practice-level maturation — a completely independent derivation context.

Question for council: Is this a legitimate additional application of Seal Four, or are we over-extending a result beyond its original scope? Does the diagonal path genuinely connect to process stage sequencing, or is the 1→2→2→1 pattern coincidental?


Claim 5: Resolution Hierarchy for Diagnostic Triage

78 positions is too much data to surface flat. We propose a resolution hierarchy — not importance ranking, but zoom level. Each layer filters for the next:

L0: Horizon balance  → Wheel vs World health across entire reading
L1: Portal vitals    → Source(10) and Creation(21) — system-wide vital signs
L2: Process check    → 22 archetypes — which processes are distorted?
L3: Bound diagnosis  → Floor/Ceiling of distorted processes — WHY?
L4: Manifest readout → Agent positions — what's actually showing up?
L5: Rebalancing path  → Vertical partner's positions — where is energy going?

Key diagnostic claim for L3: Bounds tell you WHY a process is distorted.

  • Floor distortion = process can’t start right (crooked foundation, wrong direction)
  • Ceiling distortion = process can’t complete or release (stuck at max expression)
  • Both = entire operating range compromised
  • Neither = volume problem, not a plumbing problem

Horizon consonance (visitor’s horizon × position’s horizon) computed at L0 but only surfaced when significantly skewed. Four combinations: Wheel-on-Wheel (consonant inner), World-on-World (consonant outer), Wheel-on-World (reaching), World-on-Wheel (retreating).

Spot-check performed: Drive (tooMuch) visiting Compassion’s Floor (Reverie) = “momentum blowing past the seed of connection.” Same Drive at Compassion’s Ceiling (Allure) = “momentum locked at peak magnetic pull.” Same visitor, same status, completely different diagnostic based on class — confirming L3’s value.

Question for council: Is this the right traversal order? Should bounds come before or after the process-level check? Is L0 (horizon balance) justified as a system-wide vital sign, or is it a secondary metric that belongs lower?


Claim 6: AI as Active Diagnostician (Tool-Use Architecture)

Rather than pre-computing and dumping all 78 positions into the AI prompt, we propose giving the AI diagnostic instruments via Anthropic’s tool-use API:

Seed the AI with L0-L2 (~400-500 tokens): horizon balance, portal states, process overview.

Give it 7 tools to investigate further:

  1. getHorizonBalance — system orientation
  2. getPortalState — Source/Creation deep dive
  3. getProcessDetail — single archetype detail
  4. getBoundDiagnosis — Floor/Ceiling for a process (the WHY)
  5. getManifestReadout — what’s collapsing into reality
  6. getCorrectionPath — vertical partner state + bound chain connection
  7. getHouseHealth — full practice overview with both nodes + diagonals

Each tool’s description encodes diagnostic reasoning — not just “this returns data” but “use this WHEN you see X because it reveals Y.” The tool descriptions are the AI’s medical education.

Benefits: Token-efficient (only queries what matters), scales naturally (complex readings = more investigation), the AI follows its own diagnostic hunches rather than reading pre-digested analysis.

Question for council: Does empowering the AI to investigate introduce risks? Could it follow wrong hunches and miss important signals? Should certain tools be auto-triggered rather than optional (e.g., always check rebalancing path for any distortion)?


Bonus — Trace Loop Taxonomy (Early Stage)

Trace loops (permutation cycles in the reading) have varying diagnostic complexity:

  1. Self-seated balanced — healthy, done
  2. Self-seated distorted — check rebalancing partner → may trigger loop
  3. 2-cycle, same layer — swap between two processes (productive or destructive exchange)
  4. 2-cycle, cross-layer — swap crossing signature classes (class mismatch IS the diagnostic)
  5. 3-cycle, same layer — ring of three displacements (power/friction/crisis circuit by status pattern)
  6. 3-cycle, cross-layer — loop spanning Process→Floor→Ceiling (distortion crosses resolution levels)
  7. Long cycles (4+) — systemic patterns, harder to pin down, more diffuse
  8. Nested loops sharing a position — junction point corrupting multiple circulation patterns

Each loop also has a “rebalancing shadow” — the set of vertical partners for every position in the loop. Shadow health determines whether rebalancing is available.

Question for council: Are there loop types we’ve missed? Is the complexity taxonomy useful for triage, or should the engine just report loops and let the AI reason about them?


Summary of Questions

  1. Do bound chains genuinely close within nodes, or is there a cross-node mechanism?
  2. Are the four class terms clear and honest?
  3. Is the class↔stage fractal real or decorative?
  4. Is Seal Four legitimately reappearing here, or are we over-extending?
  5. Is the resolution hierarchy the right traversal order?
  6. Does AI-as-diagnostician introduce unacceptable risks?
  7. Are there trace loop types we’ve missed?

Methodology reminder: “Derive, don’t interpret. Structure is authority.” Every claim above was derived from computational analysis of the existing architecture, not imposed from theory. If a claim breaks under scrutiny, drop it.