What Is the Slop Test in Hallmark and How Many Gates Does It Have?

The Hallmark slop test is a built-in quality-control checklist that runs 58 distinct checks—57 gates plus one pre-emit self-critique—to automatically detect AI-generated artifacts and enforce strict design constraints.

The slop test serves as Hallmark’s automated defense against synthetic-looking design patterns. As implemented in the Nutlope/hallmark repository, this verification layer inspects every generated page to ensure it meets professional standards before publication.

What Is the Slop Test in Hallmark?

The slop test is a comprehensive validation suite that identifies subtle "AI-tell" patterns—commonly referred to as slop—that make a design appear artificially produced. According to the source code in skills/hallmark/SKILL.md, the test enforces constraints such as token usage limits, color-band restrictions, and prohibitions against unwanted UI chrome. During a standard Hallmark run, the system builds the page, executes the slop-test checklist, and blocks publication if any gate returns a negative result.

How Many Gates Does the Slop Test Contain?

The slop test consists of 57 distinct gates plus an additional pre-emit self-critique, bringing the total to 58 verification points.

The 57 Core Gates

These gates represent individual binary pass/fail checks that inspect specific design attributes. As documented in skills/hallmark/references/slop-test.md, these 57 gates fire unchanged for every generation run. Each gate evaluates a unique constraint, from typography consistency to layout spacing rules.

The Pre-Emit Self-Critique (58th Check)

Beyond the 57 standard gates, the system performs a final pre-emit self-critique before outputting the finished page. This critique evaluates the overall composition using a rating system (e.g., P5 H4 E5 S4 R5 V5) rather than a simple binary result, providing a qualitative assessment of polish and professionalism.

How the Slop Test Works in Practice

When you execute a Hallmark audit, the engine automatically iterates through all 58 checkpoints. If any gate fails, the system revises the output and re-runs the test until every gate passes.

# Run the Hallmark slop test on a target file

hallmark audit path/to/page.html

A successful run produces output showing all gates cleared:

Hallmark · pre-emit critique: P5 H4 E5 S4 R5 V5
… generated HTML/CSS …
✅ Slop-test gate 01 … passed
✅ Slop-test gate 02 … passed
…
✅ Slop-test gate 57 … passed

Where the Gates Are Defined

The repository contains several key files that govern slop-test behavior:

• skills/hallmark/references/slop-test.md — Contains the definitive list of all 58 gates and their detailed definitions.
• skills/hallmark/SKILL.md — Specifies where the slop test is invoked (step 7) and how the pre-emit critique integrates with the gate system.
• README.md — Introduces the concept and explicitly states the count of "fifty-seven slop-test gates plus a pre-emit self-critique."
• site/_tests/custom/README.md — Provides practical demonstrations of the slop test applied to custom themes.

Summary

• The slop test is Hallmark’s automated quality-control mechanism that detects AI-generated artifacts.
• It contains 57 binary gates plus one pre-emit self-critique, totaling 58 checks.
• The test is invoked via the hallmark audit command and runs automatically during generation.
• All gate definitions reside in skills/hallmark/references/slop-test.md.
• Failed gates trigger automatic revision loops until the output passes all constraints.

Frequently Asked Questions

What happens if a gate fails during a Hallmark run?

If any of the 57 gates returns "no," the run immediately fails and the engine enters a revision loop. Hallmark automatically regenerates the problematic section and re-runs the slop test until every gate passes or the maximum iteration limit is reached.

Can I customize or disable specific slop-test gates?

The source code in skills/hallmark/references/slop-test.md describes the gates as firing "unchanged," indicating they are fixed constraints rather than optional settings. Users cannot selectively disable individual gates without modifying the core skill definition files.

How do I manually trigger the slop test on an existing file?

Execute the hallmark audit command followed by the target file path. This invokes the same 58-point checklist used during automatic generation, displaying individual gate results and the pre-emit critique score in the terminal output.

What is the difference between a "gate" and the "pre-emit critique"?

The 57 gates are binary pass/fail checks that verify specific technical constraints like color bands and token usage. The pre-emit self-critique is the 58th check that provides a qualitative rating across multiple dimensions (P5 H4 E5 S4 R5 V5) to assess overall design polish before final output.

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