agentleFS
2389-research2389-research/speed-run/skills/judge/SKILL.md

judge

Skill1 starsChanged 4 months agoMITView on GitHub

Scores parallel implementations with a 5-criteria worksheet (fitness, justified complexity, readability, robustness, maintainability; 25 points max), hard gates on score=1 or Fitness delta>=2, and Speed-Run efficiency tiebreakers. Invoked automatically by speed-run:showdown and speed-run:any-percent at Phase 4; not for direct user invocation.

12 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use review.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • SentrySkill

    review

    Fixture review skill.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 10 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Fitness Gate: If Fitness Δ ≥ 2 between impls → Higher fitness WINS immediately

    line 31
  2. Distinctive:

    Critical Flaw: If ANY criterion = 1 → That impl is ELIMINATED

    line 32
  3. Distinctive:

    When total scores are tied or within 1 point, prefer the implementation that:

    line 47
  4. Distinctive:

    Read: all implementation code (should already be in context)

    line 66
  5. Distinctive:

    CRITICAL: Use integer scores only (1-5).

    line 72

“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Speed-Run Judge

Score implementations using the 5-criteria framework. Fill out ALL sections exactly as shown.

Terminology: This skill uses "impl" but works for both:

  • Showdown: runner-1, runner-2, runner-3 (same design, different implementations)
  • Any%: variant-a, variant-b (different approaches/designs)

REQUIRED OUTPUT FORMAT

See references/scorecard-template.md for the complete output template. Do not summarize or abbreviate any section.

Scoring Reference

Scores Meaning

ScoreMeaning
5Excellent - exceeds expectations
4Good - fully meets requirements
3Adequate - core works, some gaps
2Poor - significant issues
1Critical flaw - disqualifying

Hard Gates (Automatic)

  1. Fitness Gate: If Fitness Δ ≥ 2 between impls → Higher fitness WINS immediately
  2. Critical Flaw: If ANY criterion = 1 → That impl is ELIMINATED

Fitness Gate Interpretation

The Fitness Gate triggers the same way in both contexts, but means different things:

ContextWhat Fitness Δ ≥ 2 Means
ShowdownOne runner deviated from or misunderstood the design. All runners should have similar Fitness since they're implementing the same spec. A large gap is a red flag.
Any%One approach genuinely solves the problem better. Different approaches can legitimately have different Fitness. A large gap means one approach is clearly superior.

In both cases, higher Fitness wins. The interpretation just explains why the gap exists.

Tiebreaker: Speed-Run Efficiency

When total scores are tied or within 1 point, prefer the implementation that:

  1. Used fewer hosted LLM fix cycles (cleaner contract prompts)
  2. Had fewer total hosted LLM calls (better task decomposition)
  3. Generated code faster (simpler, more focused prompts)

This rewards better use of the speed-run pipeline, not just code quality.

Feasibility Red Flags

Check before scoring:

  • O(n²) or worse on unbounded data
  • Unbounded memory growth
  • Self-DDoS patterns (polling, no backoff)
  • Missing pagination
  • Blocking I/O in hot path
  • No error recovery

Process

  1. Read all implementation code (should already be in context)
  2. Fill out the worksheet for EACH implementation - do not skip sections
  3. Fill out the Speed-Run Metrics table
  4. Check hard gates
  5. Announce winner with rationale (include token efficiency note)

CRITICAL: Use integer scores only (1-5). Do not use half points like 4.5.

CRITICAL: Fill out every checkbox. Do not summarize or abbreviate the worksheet.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.