CtrlK
BlogDocsLog inGet started
Tessl Logo

tooluniverse-diagnostic-test-evaluation

Diagnostic test / biomarker accuracy — sensitivity, specificity, PPV, NPV, likelihood ratios, accuracy from a 2x2 table; ROC curve, AUC, and the optimal cutoff (Youden) for a continuous biomarker; and post-test probability via Bayes. Use when you have test results vs a gold standard (binary 2x2, or a continuous score + true labels) and need to judge how good the test is, pick a threshold, or compute the probability of disease given a result. Emphasizes the prevalence-dependence of PPV/NPV.

76

Quality

95%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

SKILL.md
Quality
Evals
Security

Quality

Content

90%Weight 40%Scale 1-5

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

The body is a tight, actionable guide with copy-paste-ready commands for 2x2 metrics, ROC/AUC/Youden, and Bayesian post-test probability, plus valuable gotchas on prevalence-dependence, AUC vs operating point, and class imbalance. Minor room remains for an explicit validate-and-retry note and slightly more file-based disclosure.

Suggestions

Add a short validate/retry checkpoint after Step 2 (e.g. confirm both classes are present and the AUC CI is reasonable before quoting the Youden cutoff), even though the analysis is non-destructive.

Consider moving the metric-definition table or the Gotchas list into a short reference file and linking it from the overview to push progressive disclosure toward a clean one-level-deep split.

State explicitly that Epidemiology_bayesian post-test probability is the source of real-world PPV/NPV right inside Step 3 rather than only in the Step 1 trap note, to make the workflow self-contained at the final step.

DimensionReasoningScore

Conciseness

Lean and efficient; the metric table is framed as a prevalence-dependence reminder rather than restating basics Claude already knows, and the 'Gotchas'/'Honest limitations' sections each earn their tokens with non-obvious domain guidance.

5 / 5

Actionability

Fully executable guidance across all entry points: complete `tu run …` JSON commands, MCP `ROC_analysis(...)` call signatures with inline and CSV forms, and a real bundled script with documented CSV columns.

5 / 5

Workflow Clarity

A clear 'Which case are you in?' routing table sequences the three steps with cross-links (e.g. Step 2 feeds the chosen cutoff back to Step 1); no explicit error-recovery feedback loop, but the operations are read-only analysis so that cap is not triggered.

4 / 5

Progressive Disclosure

Well-organized overview with an accurately-referenced one-level-deep bundle file (scripts/roc_analysis.py, verified present) and clearly signaled MCP/CLI paths; most detail is inline which is appropriate for the small size, keeping it just shy of the split-file ideal.

4 / 5

Total

18

/

20

Passed

Description

100%Weight 40%Scale 1-5

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 exemplary: it enumerates concrete capabilities, supplies natural trigger phrasing for both the binary 2x2 and continuous-score cases, and explicitly states when to invoke the skill. It is comprehensive without being padded.

DimensionReasoningScore

Specificity

Lists multiple concrete actions comprehensively — sensitivity/specificity/PPV/NPV/likelihood ratios/accuracy from a 2x2, ROC curve/AUC/Youden cutoff for a continuous biomarker, and post-test probability via Bayes.

5 / 5

Completeness

Explicitly answers both what (the metrics computed) and when ('Use when you have test results vs a gold standard… and need to judge how good the test is, pick a threshold, or compute the probability of disease given a result') with concrete trigger phrases.

5 / 5

Trigger Term Quality

Comprehensive natural terms including synonyms and domain phrasing users would actually say: 'diagnostic test', 'biomarker', 'sensitivity', 'specificity', 'PPV/NPV', 'gold standard', 'ROC curve', 'AUC', 'cutoff/threshold', 'post-test probability', 'Bayes'.

5 / 5

Distinctiveness Conflict Risk

Occupies a clear niche (diagnostic test / biomarker accuracy) with distinct triggers and minimal overlap risk with adjacent statistical or epidemiological skills.

5 / 5

Total

20

/

20

Passed

Validation

100%

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

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
mims-harvard/ToolUniverse
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.