# What Is the Pre‑Emit Self‑Critique in Hallmark? Six‑Axis Quality Scoring Explained

> Understand Hallmark's pre-emit self-critique, a quality gate scoring artifacts across six axes. Learn how it ensures top quality and triggers revisions automatically for better results.

- Repository: [Hassan El Mghari/hallmark](https://github.com/Nutlope/hallmark)
- Tags: deep-dive
- Published: 2026-07-27

---

**Hallmark’s pre‑emit self‑critique is a quality gate that scores every generated artifact 1–5 across six independent axes—Philosophy, Hierarchy, Execution, Specificity, Restraint, and Variety—and automatically triggers a revision pass if any score falls below 3, ultimately stamping the final scores in a standardized file header comment.**

The `Nutlope/hallmark` repository implements this rigorous pre‑emit self‑critique to ensure that only high‑quality artifacts reach the user. Before emitting any output, Hallmark evaluates the content against six distinct quality dimensions defined in [`skills/hallmark/SKILL.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/SKILL.md). This mechanism creates a persistent feedback loop that not only enforces immediate quality standards but also allows future runs to discover and respect past critique stamps.

## The Six‑Axis Scoring System

According to the source documentation in [`skills/hallmark/SKILL.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/SKILL.md), the pre‑emit self‑critique evaluates every artifact on six independent dimensions rated from 1 (poor) to 5 (excellent):

- **Philosophy (P)** – Alignment with the brief’s conceptual intent and overall vision.
- **Hierarchy (H)** – Logical structuring and clear ordering of content elements.
- **Execution (E)** – Technical correctness, completeness, and implementation quality.
- **Specificity (S)** – Level of concrete detail and actionable guidance provided.
- **Restraint (R)** – Conciseness and disciplined avoidance of unnecessary fluff or verbosity.
- **Variety (V)** – Diversity of ideas, formats, or visual treatments presented.

## Revision Triggers and Quality Thresholds

Hallmark treats any score below 3 as a blocking defect. If **any** of the six axes receives a 1 or 2, the system automatically enters a revision pass, recomposing the artifact until all axes meet the minimum threshold of 3 or higher. This hard floor ensures consistent baseline quality before the artifact is considered complete.

## Stamping the Scores in File Headers

Once all axes pass the threshold, Hallmark records the six scores in a single‑line comment at the very top of the file. As specified in [`skills/hallmark/references/slop-test.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/references/slop-test.md), the stamp follows this exact format:

```css
/* Hallmark · pre-emit critique: P5 H4 E5 S4 R5 V5 */

```

The middot character (·) separates the system name from the critique label, and scores appear in the fixed order P, H, E, S, R, V. This standardized header serves as both a quality certificate and machine‑readable metadata.

## Discovery on Future Runs

When Hallmark processes a brief that references existing files, it scans for existing stamp comments to avoid repeating historical weaknesses. The system parses the stamped scores as a sanity check, using the previous critique to inform the current generation context. This creates a persistent quality audit trail across multiple iterations of the same project.

## Programmatic Implementation

You can identify Hallmark‑generated artifacts and verify their quality gate status by parsing the stamp comment. For example, to check if a file passed the pre‑emit self‑critique before emitting:

```javascript
function shouldEmit(content) {
  const stamp = content.match(/\/\*\s*Hallmark\s·\spre-emit\scritique:\sP(\d)\sH(\d)\sE(\d)\sS(\d)\sR(\d)\sV(\d)\s*\*\//);
  if (!stamp) return false;               // no critique → reject
  const scores = stamp.slice(1).map(Number);
  return scores.every(s => s >= 3);        // all axes meet minimum
}

```

In a generated CSS file, the stamp appears at the top:

```css
/* Hallmark · pre-emit critique: P5 H5 E5 S5 R4 V5 */
body {
  margin: 0;
  font-family: system-ui, sans-serif;
}

```

## Key Source Files

The pre‑emit self‑critique logic is defined and referenced in the following locations:

- **[`skills/hallmark/SKILL.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/SKILL.md)** – Defines the six quality axes and the revision trigger logic.
- **[`skills/hallmark/references/slop-test.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/references/slop-test.md)** – Documents the exact stamp comment syntax and parsing rules.
- **[`site/examples/wayfare/style.css`](https://github.com/Nutlope/hallmark/blob/main/site/examples/wayfare/style.css)** – Example artifact demonstrating the stamped critique comment.
- **[`site/_tests/13-alma/index.html`](https://github.com/Nutlope/hallmark/blob/main/site/_tests/13-alma/index.html)** – Test page showing the critique comment in practice.

## Summary

- Hallmark’s **pre‑emit self‑critique** scores every artifact on six axes (P, H, E, S, R, V) from 1–5 before allowing emission.
- Any score below **3** automatically triggers a revision pass until all axes meet the minimum threshold.
- Final scores are stamped in a standardized header comment: `/* Hallmark · pre-emit critique: P# H# E# S# R# V# */`.

- Future runs scan for existing stamps to avoid repeating weaknesses, creating a persistent quality audit trail.
- The implementation references [`skills/hallmark/SKILL.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/SKILL.md) for axis definitions and [`skills/hallmark/references/slop-test.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/references/slop-test.md) for stamp formatting rules.

## Frequently Asked Questions

### What does each letter in the Hallmark stamp comment represent?

Each letter corresponds to one of the six quality axes: **P**hilosophy (conceptual alignment), **H**ierarchy (structural logic), **E**xecution (technical correctness), **S**pecificity (detail level), **R**estraint (conciseness), and **V**ariety (diversity of ideas). These are scored 1–5 and appear in this fixed order after the stamp label.

### What happens if an artifact scores a 2 on one axis but 5 on all others?

Hallmark will **not** emit the artifact. Because the revision trigger is gated by the lowest score, any axis receiving below 3 forces an automatic revision pass. The system recomposes the content until all six axes achieve a minimum score of 3 or higher.

### How does Hallmark detect previous critiques in existing files?

When processing a brief, Hallmark scans file headers for the regex pattern defined in [`skills/hallmark/references/slop-test.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/references/slop-test.md), specifically looking for the comment format `/* Hallmark · pre-emit critique: ... */`. If found, the system parses the numeric scores to understand the artifact’s historical quality profile and avoid introducing previously flagged weaknesses.

### Can the pre‑emit self‑critique be disabled or customized?

According to the current source in [`skills/hallmark/SKILL.md`](https://github.com/Nutlope/hallmark/blob/main/skills/hallmark/SKILL.md), the six‑axis critique and the sub‑3 revision trigger are core quality gates built into the Hallmark skill. There is no configuration flag to disable the critique, as the stamp comment and threshold checking are integral to Hallmark’s guarantee of output quality.