CtrlK
BlogDocsLog inGet started
Tessl Logo

testland/judgment-list-author

Bootstraps human-relevance judgment lists (query sets, grading scales, rater guidelines, inter-rater agreement, Quepid tooling, TREC-style pooling, and refresh cadence) that serve as ground truth for search-relevance test suites. Use when a team needs to create or refresh the judgment corpus before running NDCG / MRR / Recall@k evaluations.

76

Quality

96%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Overview
Quality
Evals
Security
Files

Quality

Content

92%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The content is highly actionable, concise for its domain, and sequenced with strong validation feedback loops; the main gap is monolithic structure where distinct sub-procedures could be externalized into reference files.

Suggestions

Move the detailed Quepid setup workflow (Step 4) and the Cohen's kappa interpretation table (Step 5) into one-level-deep reference files (e.g. references/quepid-setup.md, references/inter-rater-agreement.md), keeping SKILL.md as a lean overview that links to them.

Consolidate the repeated inline citations of the Quepid API and TREC qrels URLs into the References section to reduce restated link text in the body.

Tighten the opening two paragraphs, which restate what the skill is for that the description and 'When to use' section already convey.

DimensionReasoningScore

Conciseness

The body is dense with domain-specific, non-obvious specifics (Cohen's kappa thresholds, TREC pool-depth guidance, Quepid export format) and avoids explaining concepts Claude already knows; quoted source material earns its place as attribution for factual claims rather than padding.

3 / 3

Actionability

Provides fully executable guidance — a runnable `cohen_kappa_score` snippet, a `pool_results` function, `docker-compose up`, concrete qrels/JSON output formats, and tables with specific numbers — matching the copy-paste-ready anchor.

3 / 3

Workflow Clarity

A clear 7-step sequence with explicit validation checkpoints and feedback loops — e.g. Step 5 gates on 'kappa < 0.60 means your guidelines are ambiguous ... re-judge the overlap set before proceeding' and Step 7 triggers re-judging at '> 30% unrated' — matching the clear-sequence-with-validation anchor.

3 / 3

Progressive Disclosure

Sections are well organized and external references are clearly signaled in a References section, but the ~290-line file is monolithic — self-contained Quepid-setup, kappa-interp1retation, and pooling material could be split into one-level-deep reference files; the skill is well above the 50-line threshold where organized sections alone would earn a 3.

2 / 3

Total

11

/

12

Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, trigger-rich, and clearly states both what it does and when to use it, with a distinct niche that avoids skill conflicts.

DimensionReasoningScore

Specificity

Lists multiple concrete artifacts/actions — 'query sets, grading scales, rater guidelines, inter-rater agreement, Quepid tooling, TREC-style pooling, and refresh cadence' — matching the anchor that enumerates several specific concrete actions.

3 / 3

Completeness

Explicitly answers both what ('Bootstraps human-relevance judgment lists ... that serve as ground truth') and when ('Use when a team needs to create or refresh the judgment corpus before running NDCG / MRR / Recall@k evaluations'), matching the both-what-and-when anchor.

3 / 3

Trigger Term Quality

Natural domain terms a search-relevance practitioner would say are well covered — 'judgment lists', 'NDCG', 'MRR', 'Recall@k', 'Quepid', 'TREC', 'search-relevance test suites' — matching the good-coverage anchor.

3 / 3

Distinctiveness Conflict Risk

A clear niche — authoring search-relevance judgment corpora — with distinct triggers unlikely to fire for unrelated skills, matching the clear-niche anchor.

3 / 3

Total

12

/

12

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Reviewed

Table of Contents