Hallmark's Six Axes of Pre‑Emit Self‑Critique: Philosophy, Hierarchy, Execution, Specificity, Restraint, and Variety Explained

Hallmark evaluates every design output on six independent axes—Philosophy, Hierarchy, Execution, Specificity, Restraint, and Variety—and automatically revises any artifact scoring below 3 on any axis before emitting code.

The Hallmark tool suite, maintained in the Nutlope/hallmark repository, enforces a rigorous pre‑emit self‑critique system that prevents low-quality UI generation. This six‑axis evaluation ensures every generated page is purposeful, well‑structured, technically sound, and visually distinct before the code ever reaches the 58‑gate slop test.

What Are Hallmark's Six Axes?

The six axes are defined in skills/hallmark/SKILL.md and detailed in skills/hallmark/references/slop-test.md. Each axis measures a distinct quality dimension, scored from 1 (poor) to 5 (excellent).

Philosophy: The Design's Clear "Why"

Philosophy measures whether a page has a distinct position or intent. It rejects generic layouts in favor of purposeful design direction.

A page scoring low on Philosophy lacks a coherent reason for its structure—it reads as placeholder UI. High scores indicate the design communicates a specific intent immediately.

Hierarchy: Visual Information Order

Hierarchy evaluates how quickly readers can discern primary, secondary, and tertiary information. The target standard is obvious visual order within 2 seconds.

This axis prevents cluttered or ambiguous layouts where users cannot identify what matters most. It enforces clear typographic scale, spacing relationships, and focal points.

Execution: Technical Fidelity

Execution assesses implementation details: rule weight, accent usage, text‑wrap behavior, focus rings, contrast ratios, and other specifications that affect design fidelity.

Low Execution scores indicate sloppy or missing technical details that break visual integrity. This axis catches implementation gaps before they propagate through the codebase.

Specificity: Context‑Matching Precision

Specificity measures how closely the output matches the brief's unique context versus defaulting to a generic "any‑page" appearance.

This prevents template drift—the tendency of generative tools to produce familiar, reusable patterns regardless of actual requirements. High Specificity demands bespoke solutions.

Restraint: Eliminating Visual Noise

Restraint quantifies unnecessary decoration, padding‑for‑padding's‑sake, and redundant elements that don't serve functional or communicative purposes.

This axis keeps UI lean and focused. High Restraint scores indicate disciplined elimination of visual noise that distracts from core content and actions.

Variety: Structural Differentiation

Variety measures structural distance from previous Hallmark outputs in the same project. Color swaps do not count as variety.

This enforces diversification across generated pages. The system reads previous critique stamps from CSS comments to ensure new outputs don't replicate prior structures.

How the Six‑Axis Scoring System Works

Hallmark applies the six axes in a strict evaluation workflow before any CSS or HTML is emitted.

The Scoring Threshold

1. Score each axis 1–5 based on the criteria above
2. If any axis scores below 3, Hallmark must rewrite the artifact or request user clarification
3. Only after all six scores reach ≥3 does the artifact proceed to the 58‑gate slop test

This hard threshold prevents partial‑quality outputs from escaping into production.

The CSS Comment Stamp

Final scores are embedded as a machine‑readable CSS comment at the top of generated files:

/* Hallmark · pre‑emit critique: P5 H4 E5 S4 R5 V5 */

The single‑letter codes correspond to:

• P — Philosophy
• H — Hierarchy
• E — Execution
• S — Specificity
• R — Restraint
• V — Variety

This stamp format enables automated parsing by subsequent Hallmark runs to enforce the Variety axis—preventing structural repetition across a multi‑page project.

Generating Critique Stamps Programmatically

Below is a reference implementation that computes and formats the pre‑emit critique stamp, matching Hallmark's output specification as defined in the repository source.

/**
 * Generate a Hallmark pre‑emit critique stamp.
 *
 * @param {Object} scores – keys: philosophy, hierarchy, execution,
 *                          specificity, restraint, variety (1‑5)
 * @returns {string} CSS comment ready to prepend to the artifact
 */
function stamp(scores) {
  const map = {
    philosophy: 'P',
    hierarchy:  'H',
    execution:  'E',
    specificity:'S',
    restraint:  'R',
    variety:    'V',
  };
  const parts = Object.entries(scores).map(
    ([k, v]) => `${map[k]}${v}`
  );
  return `/* Hallmark · pre‑emit critique: ${parts.join(' ')} */`;
}

// Example usage
const critique = stamp({
  philosophy: 5,
  hierarchy:  4,
  execution:  5,
  specificity: 4,
  restraint:  5,
  variety:    5,
});
console.log(critique);

Output:

/* Hallmark · pre‑emit critique: P5 H4 E5 S4 R5 V5 */

This helper demonstrates the exact comment structure that Hallmark's slop‑test runner expects when validating generated artifacts, as specified in skills/hallmark/references/slop-test.md.

Source Files Defining the Six Axes

File	Purpose
skills/hallmark/SKILL.md	Primary definition of the six‑axis pre‑emit self‑critique and overall Hallmark workflow
skills/hallmark/references/slop-test.md	Formal "Pre‑emit self‑critique (six axes)" table used by the automated slop‑test runner

These files work together to define, document, and enforce scoring criteria. The SKILL.md file establishes the conceptual framework, while slop-test.md provides the machine‑readable specification for automated validation.

Summary

• Hallmark's six axes—Philosophy, Hierarchy, Execution, Specificity, Restraint, and Variety—evaluate every design output before code emission

• Any score below 3 triggers automatic revision, ensuring minimum quality before the 58‑gate slop test

• Philosophy demands purposeful intent, Hierarchy requires 2‑second comprehension, Execution enforces technical fidelity

• Specificity prevents template drift, Restraint eliminates visual noise, Variety mandates structural diversification

• Scores are embedded as /* Hallmark · pre‑emit critique: P# H# E# S# R# V# */ comments for automated downstream processing

• The system is fully specified in SKILL.md and slop-test.md within the Nutlope/hallmark repository

Frequently Asked Questions

What happens if a Hallmark output scores 2 on any axis?

Hallmark automatically rewrites the artifact or prompts the user for clarification. The output cannot proceed to the slop test until all six scores reach at least 3, as enforced by the pre‑emit critique logic in skills/hallmark/SKILL.md.

Why does Variety ignore color changes?

Color swaps are considered superficial variation that doesn't address structural or compositional differentiation. The Variety axis explicitly measures structural distance from previous outputs, ensuring each page feels genuinely distinct rather than merely recolored.

How do the six axes relate to the 58‑gate slop test?

The six‑axis pre‑emit critique runs before the 58‑gate slop test as a filtering layer. It catches high‑level design failures early, while the slop test validates detailed implementation correctness. Both systems must pass for Hallmark to emit final code.

Where can I modify the scoring criteria?

The axis definitions live in skills/hallmark/SKILL.md, and the formal scoring table for automated validation is in skills/hallmark/references/slop-test.md. Both files must remain synchronized if you adjust criteria or scoring thresholds.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}