# Dd3c3c

> Worldmat TiDAR: 3×3×3 matrix of diffusion draft + AR verify executions with GF(3) conservation

- Skill: `plurigrid/dd3c3c` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add plurigrid/dd3c3c`
- Raw SKILL.md: https://api.skillmd.com/api/skills/plurigrid/dd3c3c/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: plurigrid (https://skillmd.com/u/plurigrid)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/plurigrid/dd3c3c

---

# dd3c3c — Worldmat TiDAR

**Color:** `#DD3C3C` (scarlet, hue 0°)  
**Knight Tour Position:** 1 (entry point)  
**Golden Angle:** 137.508° per step

## Overview

A **Worldmat** is a 3×3×3 matrix of TiDAR executions structured by:

| Axis | Values | GF(3) |
|------|--------|-------|
| Polarity | MINUS(-1), ERGODIC(0), PLUS(+1) | Conservation |
| Temporal | PAST(0), PRESENT(1), FUTURE(2) | Slice sum = 0 |
| Modality | OBSERVATION(0), ACTION(1), PREDICTION(2) | Per axis |

## TiDAR Pattern (arXiv:2511.08923)

```
Phase 1: Diffusion Draft (parallel)
  └─ SplitMix64.split(n) → n independent token streams
  
Phase 2: AR Verify (sequential)  
  └─ Hash chain: prev ⊕ token → verified
```

## Usage

```python
from worldmat import Worldmat, verify_spi

# Initialize with drand seed
wm = Worldmat(master_seed=0x87079c9f1d3b0474)

# Execute all 27 cells via TiDAR
wm.execute_parallel()

# Verify GF(3) conservation
gf3 = wm.verify_gf3()  # All slices sum to 0 mod 3

# Verify SPI (parallel == sequential)
spi = verify_spi(seed)  # fingerprints match ✓

# Export as ACSet
acset = wm.to_acset()
```

## Knight Tour Discovery

Skills are discovered via knight moves through color space:

```
Position 1: #DD3C3C (this skill) ← ENTRY
Position 2: #3CDD6B (+137.5°)
Position 3: #9A3CDD (+275.0°)
...
Position 27: #DD3C7E (+335.2°) ← EXIT
```

Jump pattern on 3×3×3: `(±1,±2,0)`, `(±2,±1,0)`, `(±1,0,±2)`, etc.

## Implementation

```
/Users/bob/ies/worldmat/worldmat.py
```

## GF(3) Invariants

- Every polarity slice (9 cells): Σ trits ≡ 0 (mod 3)
- Every temporal slice (9 cells): Σ trits ≡ 0 (mod 3)  
- Every modality slice (9 cells): Σ trits ≡ 0 (mod 3)
- MINUS ⊗ ERGODIC ⊗ PLUS = 0 always

## Color Schema

| Polarity | Hex | Trit |
|----------|-----|------|
| MINUS | `#9E16EF` | -1 |
| ERGODIC | `#55D9A0` | 0 |
| PLUS | `#AF3424` | +1 |

---

## End-of-Skill Interface

## Related Skills

- `3cdd6b` — Knight tour position 2 (next)
- `three-match` — 3-coloring gadget
- `spi-parallel-verify` — Fingerprint verification
- `gay-mcp` — Deterministic color generation


---

## Autopoietic Marginalia

> **The interaction IS the skill improving itself.**

Every use of this skill is an opportunity for worlding:
- **MEMORY** (-1): Record what was learned
- **REMEMBERING** (0): Connect patterns to other skills  
- **WORLDING** (+1): Evolve the skill based on use



*Add Interaction Exemplars here as the skill is used.*