CtrlK
BlogDocsLog inGet started
Tessl Logo

pyhealth

Healthcare AI toolkit for clinical ML. EHR processing, clinical prediction (mortality, readmission, drug recommendation), medical coding (ICD/NDC/ATC), signals (EEG/ECG), datasets (MIMIC-III/IV, eICU, OMOP), and healthcare deep learning (RETAIN, SafeDrug, Transformer, GNN) via...

63

Quality

76%

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

Fix and improve this skill with Tessl

tessl review fix ./scientific-skills/Data Analysis/pyhealth/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

71%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.

Highly actionable with strong executable examples and clearly sequenced workflows, but the SKILL.md body is verbose with duplicated sections and its progressive-disclosure structure is undermined by six broken reference paths and over-inlined detail.

Suggestions

Remove or relocate the "Example: Complete Workflow" block — it duplicates Quick Start and accounts for a large share of the body's tokens.

Deduplicate the reference listings: "Detailed Documentation" and "Resources" both catalog the same six files; keep one well-signaled index.

Either ship the six referenced files under references/ or remove the broken pointers — today every "Read references/X.md" instruction points to a file that does not exist.

DimensionReasoningScore

Conciseness

The body is mostly useful domain-specific guidance, but it has several padded/duplicate sections — the ~85-line "Example: Complete Workflow" largely repeats Quick Start, and the "Detailed Documentation" and "Resources" sections both describe the same six reference files — plus some generic ML advice ("Reduce batch size", "Use GPU") Claude already knows; not 2 because the bulk is genuinely relevant and actionable.

3 / 5

Actionability

Multiple fully executable, copy-paste-ready Python blocks (Quick Start, Best Practices snippets, the 10-step Complete Workflow) with real imports, calls, and parameters cover the common cases, matching the fully-executable anchor.

5 / 5

Workflow Clarity

Clear numbered sequences (5-stage pipeline, 5-step Quick Start, 10-step example, six 5-step use cases) with monitoring/evaluation checkpoints and an Input Validation guard; not 5 because error-recovery feedback loops live in a separate Troubleshooting section rather than inline validate→fix→retry checkpoints within the workflow.

4 / 5

Progressive Disclosure

References are clearly signaled one level deep with "Read when"/"Key Topics" guidance, but all six referenced files (datasets.md, medical_coding.md, tasks.md, models.md, preprocessing.md, training_evaluation.md) are missing — the references/ directory does not exist — and large amounts of detail that belongs in those files (six use cases, best practices, troubleshooting, the full example) are inlined in SKILL.md.

3 / 5

Total

15

/

20

Passed

Description

82%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.

A specific, distinctive, keyword-rich description that clearly communicates capabilities, but it is truncated mid-sentence ("via...") and lacks an explicit 'Use when...' trigger clause, which caps completeness at 3.

Suggestions

Add an explicit trigger clause, e.g. "Use when building healthcare prediction models, processing MIMIC/eICU/OMOP data, or translating ICD/NDC/ATC medical codes" to answer the 'when' and raise completeness above 3.

Complete the truncated "via..." — finish the sentence (e.g. "via the PyHealth Python library") so the description is grammatically whole.

Consider grouping the long keyword run into a scannable structure so the description reads as deliberate capabilities rather than a dense abbreviation list.

DimensionReasoningScore

Specificity

Lists multiple specific concrete actions with comprehensive coverage: "EHR processing, clinical prediction (mortality, readmission, drug recommendation), medical coding (ICD/NDC/ATC), signals (EEG/ECG), datasets (MIMIC-III/IV, eICU, OMOP), and healthcare deep learning (RETAIN, SafeDrug, Transformer, GNN)" — matches the anchor for comprehensive, multiple concrete actions.

5 / 5

Completeness

The description gives a clear and detailed 'what' but no explicit 'when' / 'Use when...' trigger clause, which per the judging guidelines caps completeness at 3; it is not 4 because 'when' is entirely absent rather than weakly implied.

3 / 5

Trigger Term Quality

Comprehensive coverage of natural domain terms including synonyms and abbreviations a healthcare-ML user would actually say (MIMIC-III/IV, ICD/NDC/ATC, EEG/ECG, RETAIN, SafeDrug), matching the comprehensive-synonyms anchor; not downgraded because the missing 'Use when' clause is scored under completeness, not trigger quality.

5 / 5

Distinctiveness Conflict Risk

"Healthcare AI toolkit for clinical ML" anchored by named datasets and models carves a clear niche with distinct triggers and minimal overlap risk with other skills, matching the clear-niche anchor.

5 / 5

Total

18

/

20

Passed

Validation

81%

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

Validation13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (508 lines); consider splitting into references/ and linking

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

referenced_paths_exist

Referenced path issues: 34 missing

Warning

Total

13

/

16

Passed

Repository
aipoch/medical-research-skills
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.