# toroid.fyi — cross-instrument review, 2026-10-02

Provenance: Claude-generated (Fable 5.1), at the analyst's request, from a line-by-line read of the
repo at commit 1608be9. Everything below is proposal until the analyst ratifies it. Counts marked
**verified** were computed from the data files by script (`cycles/build/audit.mjs`); everything
else is reading, inference, or synthesis and is labelled as such.

---

## Part 1 — Macro and meta analysis

### 1.1 One loop, described four times, at three resolutions (verified structure; synthesis of the equivalence)

The site carries four descriptions of a closed extraction cycle, each at a different station count:

| Instrument | Stations | Order |
|---|---|---|
| Polymorphic Matrix (2026-03) | 4 | apparatus → convergence (MIF) → erosion → feedback |
| Substrate Attack (2026-09) | 3 | cut the layer → frame the cut → collect on the damage |
| Removal Economy (2026-05) | 6 | substrate → threat → removal → harm → extraction → cost of lives |
| Deception Arcane chains (v0.3+) | 3 | S3 block the fix → sustained crisis → S6 capture |

These are the same operator sampled at different rates. The common six-phase form, into which all
four map without remainder:

1. **Substrate** — the load-bearing layer in place (Matrix: the polity the apparatus acts on; Removal I)
2. **Framing** — the layer recoded as threat, waste, or incompetence (Removal II; Substrate "frame the cut"; Arcane S5)
3. **Cut** — the operation (Matrix apparatus; Removal III; Substrate "cut"; Arcane S3 refusal of the fix; every 7.x cell)
4. **Harm** — the material residue (Matrix erosion; Removal IV; the register's baselines)
5. **Capture** — value re-booked privately (Removal V; Substrate "collect"; Arcane S1/S6)
6. **Reinvestment** — harm converted into the justification and the appropriation for the next cut (Matrix feedback; Removal "the harm of each turn financing the next"; the ratchet doctrines)

The phase the site under-measures is **6**. Every instrument asserts that harm re-enters as
justification; none measures the conversion (harm → appropriation → next cut) as a quantity. The
Removal page states it ("$170B booked as a loss … routed into Station V") but it is the only place.

### 1.2 The second periodicity the torus promised and the instruments never used (synthesis)

Every page says the torus is "doubly periodic". Only the Matrix uses two circles (major: the
five-vector pentagon; minor: apparatus → MIF → erosion → feedback). The Catalysis Toroid nests two
tori instead; the Removal loop is one circle of six; the opening ring is one circle of seven. The
minor circle has been decoration since March.

The site's own method supplies the missing cycle. Read across pages, the work follows a closed
loop of its own:

**observe → name → measure → predict → check → correct → observe**

- observe: the Catalysis Toroid cells, the register's evidence annex
- name: the predicates, the Matrix (the first naming), the Substrate Attack page, the book
- measure: the Deception Arcane signatures and composites
- predict: the register's locked predictions, Sches as the commitment model
- check: dated checkpoints, negative controls, adversarial tests T1/T2, the pool-footage baseline
- correct: `_revision_note`, withdrawn claims (the 70/85 count justification), the ρ −0.89 → −0.57 de-bias

That is a second independent cycle. Two independent cycles is exactly the condition under which a
torus is the right 3-manifold boundary and not an ornament: the surface is the product of the two
circles, and every point on it is a pairing (phase of the extraction cycle, phase of the
measurement cycle).

The payoff is that obfuscation gets a geometric definition. The Substrate Attack page already says
the measuring apparatus (courts, inspectors general, audits, press) is itself load-bearing and is
being cut. On the two-cycle surface that is literal: **the regime's cycle runs around the major
circle; the observer's cycle runs around the minor; obfuscation is the regime's work to keep the
minor circle from closing** — IG firings cut *observe*, manufactured incoherence (S5) cuts *name*,
anti-remediation (S3) cuts *correct*, the press ban cuts *observe* and *check* at once. The
register's PRESS-ACCESS prediction is therefore the first instrument on the site pointed at the
minor circle itself: measurement of the attack on measurement. That reflexive position (observer
inside the observed system, second-order cybernetics — von Foerster's term, offered as a link not
a citation) is the site's actual meta-claim and nothing currently draws it.

### 1.3 Each instrument is a different calculus operation on the same field (synthesis)

| Instrument | Operation |
|---|---|
| Catalysis Toroid | the state — who acts on whom, position on the surface |
| Polymorphic Matrix | the flow field — direction of circulation |
| Deception Arcane composite | the level — a scalar reading of a state |
| Arcane temporal layer | the first derivative (velocity), the count-derivative (breadth), the timing correlation (synchrony) |
| Register | the forward integral — a prediction is a stated future value with a check date |
| Removal Economy | one trajectory — a single particle traced around the loop |
| Sches | the generator — how agents committing blind produce the loop at all |
| Substrate Attack | the invariant — the conserved quantity being moved |
| Named and Counted | the units — what gets counted and in what |

Nothing on the site states this division of labour. Visitors are told the instruments "share a
topology"; they are not told that they are complementary operations and that the set is nearly
complete. "Nearly": there is no second-derivative (acceleration of capture) and no instrument for
phase 6.

### 1.4 The conserved quantity, and a cross-disciplinary reading (synthesis, flagged as analogy)

The Substrate Attack page names the single currency: the financial health of the nation. The
Removal page quantifies one turn ($96.7B paid in; $170B appropriated; $8.9B/yr destroyed per
million removed; 31 deaths off-ledger). Put together, the site is describing a **transfer, not a
destruction**, with a loss term: value moves from the commons to private capture; what cannot be
transferred (lives, separations) is the loss term the contracts keep off the books.

Thermodynamic analogy, offered as a thinking tool, not a claim: a heat engine moves energy across a
gradient and does work; the insulation is what stops the back-reaction. Here the gradient is
commons → capture; the work is extraction; the insulation is obfuscation; the waste heat is the
unpriced harm. The Arcane's five signatures then read as engine parameters — S1 the gradient
(stated vs realized beneficiary), S2 selective conductivity, S3 a closed relief valve, S4 the
material the engine is built from, S5 the insulation. That S5 carries the highest weight (0.30) is
consistent with the analogy: without insulation the engine stalls. This is why the analogy is
worth having on the page as a reading aid and nowhere as a finding.

### 1.5 The vortices and the vectors are the same six-and-five, never mapped (verified counts; mapping is inference)

| Catalysis vortex (6) | Matrix pressure → erosion (5) |
|---|---|
| capital, money | economic → democratic erosion |
| attention, narrative | cognitive → epistemic collapse |
| power | institutional → institutional decay |
| intelligence | covert → sovereignty transfer |
| — | legitimacy → social fracture |

Finance is split in two on the toroid and nothing carries legitimacy/identity consolidation. Either
the vortex set is missing a sixth-equivalent (a "faith/identity consolidation" vortex — the cells
exist: Faithocracy 38, Ethnocracy 5, Xenocracy 8, christian_nat and hindutva actors) or the Matrix
over-counts. Decision for the analyst; the diagram shows the hole.

### 1.6 The largest evidence complex has no measurement (verified)

Cells by complex: Tech-media **150**, GovMilitary 123, PetroMedical 118, Prison & Labor 56, Sugar &
Food 36. The Arcane's seven architectures are all fiscal/service (GovMilitary, PetroMedical,
Prison & Labor). Tech-media — the largest complex — has zero scored architectures; the register's
press-access prediction is the first probe. Candidate architecture for the slate: the
public-information apparatus (T1 is already pre-declared for it).

### 1.7 Scale ladder (synthesis)

cell (one actor on one target) → architecture (a program, scored) → regime (the mode mix, the
vortices) → condition (the substrate attack) → and back: the condition selects which cells are
cheap to open next. That return edge is phase 6 again. The ladder is the minor circle at a
different grain, which is the reason the two-cycle model, and not a three-cycle one, is enough: the
scale ladder and the method cycle move together.

---

## Part 2 — Audit: missing connections, false connections, errors, improvements

Severity: **A** false or broken in what the visitor sees · **B** inconsistency in stated facts ·
**C** design / UX · **D** upgrade.

### A — false or broken

A1 **Dual-role ids break actor labels (verified, 12 nodes).** Thirteen ids exist as both an actor
and a target (`doj, state_dept, congress_maj, iran_post, foundation_models, big_tech, cartels, ftc,
fcc, hhs, civil_society, public_sector_workers, scotus`). `nodeLabels` is keyed by bare id, so the
target's label overwrites the actor's; the actor spheres for Big Tech, Congress, DOJ, FTC, FCC,
HHS, foundation models, civil society and the public-sector workforce can never show a label
(permanently `display:none` from init). Fix: key labels by role (`a:`/`t:`). The adjacency merge of
dual-role ids is correct behaviour (one entity, two roles) and is kept.

A2 **Relationship layers are 39% invisible (verified).** 30 actors have no 7.x cell and are
filtered out of the rendered set; every R/B/P edge touching them is silently skipped. Not drawn:
**51 of 130 R-cells, 9 of 50 B-wormholes, 4 of 17 P-cells.** Among the missing: R038 (Executive →
Federalist Society staffing), which the Deception Arcane glossary cites by id as its S4 anchor. The
evidence layer cannot show the cell the measurement layer points to. Fix: render any actor that has
at least one edge of any layer; show the honest count where the visitor can see it.

A3 **Vortex source that does not exist (verified).** `capital` vortex lists `sovereign_wealth`;
no such actor. The sovereign-wealth actors are `saudi`, `uae`, `qatar` (State Allies). Fix: point
at the actors that exist (`saudi`, `uae`; `qatar` has no cells and would still not draw until A2).

A4 **`scotus` missing from `actorFlow` (verified).** Only actor without a hemisphere assignment;
falls through to whatever the placement code defaults to. Fix: assign (`upper` — direction-source
by analogy with the judiciary cluster's role in the cells).

### B — stated facts that disagree with the data (all verified)

B1 Actors: README, /about, /tour and the toroid onboarding say **136**; the data has **137**
(`scotus` was added as an actor without the texts following).

B2 Matrix: /tour says **200 edges, 8 feedback loops**; the data has **195 edges, 11 feedback edges**.

B3 Multimalocracy: the /predicates/multimalocracy page enumerates 42 slots (40 distinct names
once the four cross-cluster duplicates and the two matrix-derived additions are counted). The
toroid's cells use **three modes that are on no slot** — Necrocracy (73 cells), Kleptocracy (33),
Theocracy (1) — and **ten listed modes are used by no cell** (Anemocracy, Bureaucratocracy,
Complexocracy, Consumerocracy, Cynicocracy, Diplocracy, Narratocracy, Oligarchy, Polygarkracy,
Ridiculocracy). The methodology paper names Kleptocracy and Necrocracy as members, so the slot
page is the stale artifact. Sabbotarchy itself is deliberately not a slot ("operates behind the
other modes"), which is consistent. One tension stands: slot 14 is **Obfuscracy** ("intentional
complexity"), and seven cells carry it as a mode, while predicate 002 says Obfuscratic is a
modifier and not a regime type. Decision for the analyst (see questions).

B4 Deception Arcane status paragraph says T2 "returned 48"; the table, the JSON and the register
say **50** after the 2026-09-21 S2 re-score. Both are true at their dates; the sentence should say so.

B5 Tier calibration: 288 Documented / 57 Inferable / **4 Speculative**. Not an error; recorded
because a 1% speculative rate on 349 interpretive cells is itself a claim the audit should see.

### C — design and UX

C1 Opening ring order is roughly chronological (poly, sabbot, removal, book, substrate, sches,
matrix) and carries no meaning. With the two-cycle model the ring can be ordered as the measurement
cycle (observe → name → measure → predict → check → correct), so the first thing a visitor sees is
the method, and the new gate sits where it belongs.

C2 Opening ring has eight gates after the addition; label width 210px at radius 2.5 is tight at
the 4 and 8 o'clock positions on 1280-wide screens. Radius 2.65 and camera z 8.2 keep the labels
clear.

C3 The ring is a set of toroids with no connective tissue. A faint closed line through the gate
centres makes the ring read as a cycle rather than a menu, at near-zero visual cost.

C4 Site-wide: the Removal page footer promises "integration with the Catalysis Toroid cell
references (B040, 7.277, 7.328)" as future work. The cells exist. The toroid has no deep-link
parameter, so the promise cannot be kept without one.

C5 New page must carry the same no-JS static index and reduced-motion path as `/index.html`.

### D — upgrades the review found

D1 Deep links into the toroid (`/poly/?cell=B040`, `?actor=ice_dhs`) — the single missing
connection that would let the Arcane's anchors, the Removal footnotes and the new page land on the
exact evidence. Modest change: read the parameter after init, run the existing selection.

D2 A "reinvestment" quantity for phase 6: the Removal page's own numbers (per-turn appropriation vs
per-turn revenue destroyed) are a first instance; the Arcane's S6 selective-allocation is the
general form. Proposal only.

D3 Candidate seventh vortex (legitimacy/identity consolidation) — §1.5. Proposal only.

D4 Candidate eighth architecture: public-information apparatus (Tech-media complex) — §1.6, T1.
Proposal only.

D5 A dated field on 7.x cells. The Arcane has epochs; the register has dates; 349 cells have none
(93 mention a 2023–2026 year in prose). A unified time axis is the precondition for the
acceleration measurement in §1.3. Proposal only.

---

## Part 3 — What was built in this pass (branch `two-cycles` on the mini, not merged, not pushed)

Commit 560acb4, verified in a headless browser (desktop 1400px, phone 390px; no console errors):

- `/cycles/` — the two-cycle meta-instrument: 3D torus (major = extraction cycle, minor =
  measurement cycle), instruments drawn as patches on the surface, gaps outlined, the unrolled
  fundamental domain as the legible 2D view (hover-linked to the form), equivalence table,
  operations table, the cuts on the measurement cycle, the engine analogy as a reading aid, a
  falsification paragraph, provenance.
- `/cycles/build/audit.mjs` — reproduces every count in this review from the data files.
- `/index.html` — eighth gate; ring reordered to run as the measurement cycle from the top (C1);
  radius 2.65 / camera retuned for eight (C2); connective ring line (C3); description, intro
  and static index updated.
- `/poly/` — A1 role-keyed labels; A2 render edge-bearing actors (137/137 drawn, 130 R · 50 B ·
  17 P all with endpoints) and a stats line that says so; A3 `sovereign_wealth` → `saudi`, `uae`;
  A4 `scotus` flow entry; D1 deep links `?node=` / `?cell=` (tested: R038 → links layer on,
  7.297 → its channel on and the edge drawn, B040 → wormholes on).
- Removal footnotes and the Arcane glossary link into the deep links (C4).
- B1, B2, B4 corrected in prose. B3 left in the data; the taxonomy page carries the reconciliation
  note.

### Second pass — found while checking the list

- `/about/` and `/tour/` still say "three instruments"; the surface now says nine forms. Stale
  framing, not wrong facts — the tour is a dated walkthrough and can say so. Not changed.
- `/tour/` chapter IV counts "130 R-cells · 50 B-wormholes · 17 P-cells" as rendered; after A2
  that is now true. Before A2 it was not.
- The Catalysis Toroid's onboarding tip 1 says actors "will appear" while actor spheres are
  already visible at load. Pre-existing copy, harmless; noted.
- The site has no "what changed" entry for this pass yet; `/updates/` should carry one when
  this merges (drafted below the questions in the hand-off, not written into the page).
- Phase-6 measurement, the seventh vortex, the eighth architecture and dated cells (D2–D5) remain
  proposals; none was built, by design — each is the analyst's call.

### Third pass — the documents (side job, 2026-10-02)

- Every `.md` on the site was served as plain text, so papers read as raw markup. Added
  `tools/render-md.mjs` (vendored marked, MIT) which renders each document to a sibling `.html`
  reader page (`assets/reader.css`), keeps the `.md` as downloadable source, and relinked all
  37 documents across 11 pages. Re-run the script after editing any `.md`.
- `/tour/` had relative links and iframe sources (`./poly/`, `./matrix/` …) that resolved to
  `/tour/poly/` — 404 on the live site, confirmed. All four embeds and the document links were
  broken since the page moved under `/tour/`. Fixed to absolute paths.
- The node reference carried a pasted duplicate of three paragraphs and the phrase "the framework
  *you originated*" — assistant voice that had acquired the author's byline. Duplicate removed,
  phrase cut to "the framework", in both copies (`poly/` and `papers/`). Flagged under the
  provenance rule; no other second-person text found in the documents.

### Open questions for the analyst (nothing ships until these are answered or waived)

1. **Name.** "The Two Cycles" is descriptive and provisional; the gate, title and file path all
   change if a coined name is wanted.
2. **Coverage weights.** The 0 / thin / strong readings per instrument are the drafter's; the
   grid is only as good as they are. Any cell he reads differently moves on the torus.
3. **B3.** Reconcile the taxonomy by (a) adding Necrocracy, Kleptocracy, Theocracy as slots,
   (b) remapping the 107 cells onto existing slots, or (c) leaving the page as a 42-slot
   historical artifact with the note. And whether "Obfuscracy" (slot 14, 7 cells) survives
   predicate 002.
4. **Ring order.** Reordering the opening ring is a design change to the first thing every
   visitor sees. It can be reverted independently of everything else.
5. **A2 side-effect.** Thirty more actor spheres on the outer torus; most sit in the "boundary"
   band. Density is up; the hub-label rule keeps the names quiet. Worth a look on the mini.

### Resolutions (2026-10-02, on the analyst's instruction to correct and actualize)

- **Q3, taxonomy.** Necrocracy and Kleptocracy added as proposed additions 45 and 46 beside the
  two matrix-derived ones (the "42 numbered slots" claim stands); the one Theocracy tag folded
  into Faithocracy on cell 7.52. Slot 14 (Obfuscracy, a mode) and predicate 002 (Obfuscratic, a
  modifier) survive together once glossed: a regime can be Obfuscratic without being an
  Obfuscracy. The gloss is on the slot.
- **Q4, ring order.** Was: Poly, Sabbot, Removal, Book, Substrate, Sches, Matrix, starting at the
  bottom. Is: Poly, Matrix, Book, Substrate, Removal, Sabbot, Sches, Two Cycles, starting at the
  top, clockwise — the measurement cycle. Reverting means restoring the earlier `RING` array and
  `ringPos` sign in `index.html`; nothing else depends on the order.
- **Reorganisation items 1–6** applied: papers canonical under `/papers/` with 301s from the old
  instrument-folder copies; Catalysis Toroid and Deception Arcane open dark; `landing/` retired
  (301 → `/`); `/about/` rewritten to the nine forms and the tour date-stamped; The System lists
  everything beneath the surface; an `/updates/` entry for the release.
- Q1 (name), Q2 (coverage weights) and Q5 (the thirty new spheres) remain open; shipped as
  provisional and labelled so on the page.
