experiment-design

Design scientific experiments including sample size calculation, randomization, control groups, blinding, and study protocols. Covers RCTs, quasi-experiments, factorial designs, A/B tests, survey design, and observational studies. Use when user asks to design an experiment, ca...

Quality

70%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./scientific-skills/Protocol Design/experiment-design/SKILL.md

Quality

Content

61%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.

A genuinely actionable, well-structured reference (executable power-analysis code, decision table, protocol template) held back by a leaked personal environment path, some over-explanation of basic methodology concepts, and the absence of explicit validation checkpoints in the workflow.

Suggestions

Remove the hardcoded "source /Users/zhangmingda/clawd/.venv/bin/activate" line (a personal environment leak) so the power-analysis code is portable and copy-paste ready; assume scipy is available or state the dependency plainly.

Tighten "Key Design Principles" — Claude already knows what positive/negative controls and single/double/triple blinding are; keep only the skill-specific guidance (e.g., when to choose block vs. stratified randomization) and drop the textbook definitions.

Add explicit validation checkpoints to the protocol workflow, e.g. after sample-size calculation verify power ≥ 0.80 and after drafting the protocol confirm all 12 sections are populated before declaring it complete.

Dimension	Reasoning	Score
Conciseness	Mostly efficient tables and code, but the "Key Design Principles" section restates concepts Claude already knows (definitions of positive/negative controls, single/double/triple blinding, simple/block/stratified/cluster randomization), and the power-analysis block leaks a non-portable personal path ("source /Users/zhangmingda/clawd/.venv/bin/activate").	3 / 5
Actionability	Provides copy-paste-ready Python (sample_size_ttest, sample_size_chi2, sample_size_correlation), a concrete protocol template, and a decision table, but the hardcoded "source /Users/zhangmingda/..." activation line breaks out-of-the-box executability for anyone but the author.	4 / 5
Workflow Clarity	The protocol template gives a sequenced 12-section structure and "Error Handling" offers generic guidance, but there are no explicit validation checkpoints or a validate→fix→retry loop integrating power checks, randomization verification, or protocol-completeness checks into the workflow.	3 / 5
Progressive Disclosure	Well-organized into clearly headed sections (Design Selection Guide, Power Analysis, Key Design Principles, Protocol Template, Pre-registration, Tips) with no nested references and no bundle files present, but it is a single monolithic file rather than an overview pointing to one-level-deep reference files.	4 / 5
	Total	14 / 20 Passed

Description

80%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, well-scoped description with strong capability coverage and a clear use-trigger, weakened by a truncated "Use when" clause that ends mid-word ("ca...") and a few missing common trigger terms.

Suggestions

Complete the truncated "Use when user asks to design an experiment, ca..." clause — finish the sentence and add concrete secondary triggers (e.g. "calculate sample size", "plan a clinical trial", "choose a study design").

Add high-frequency natural terms users actually say — "power analysis", "clinical trial", "study design" — to improve trigger-term coverage toward the comprehensive anchor.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — "sample size calculation, randomization, control groups, blinding, and study protocols" — plus six named design types (RCTs, quasi-experiments, factorial designs, A/B tests, survey design, observational studies), matching the comprehensive-coverage anchor.	5 / 5
Completeness	Both "what" and "when" are present ("Use when user asks to design an experiment"), but the when-clause is a single brief trigger that is cut off mid-word ("ca..."), so it could be more explicit and complete rather than matching the clear-concrete-trigger-phrases anchor.	4 / 5
Trigger Term Quality	Good natural-term coverage ("design an experiment", "sample size", "RCT", "A/B test", "survey") with synonyms, but common user phrases like "power analysis", "clinical trial", and "study design" are absent, so it falls short of the comprehensive anchor.	4 / 5
Distinctiveness Conflict Risk	The experiment-design niche (randomization, blinding, RCTs, protocols) is mostly distinct with specific triggers, with only minor overlap risk against generic statistics or A/B-testing skills, fitting the mostly-distinct anchor.	4 / 5
	Total	17 / 20 Passed

Validation

93%

Warnings & errors only

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

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: aipoch/medical-research-skills
Path: scientific-skills/Protocol Design/experiment-design/SKILL.md
Commit: 63c61d3

Reviewed: 5 hours ago

Table of Contents

Discovery Implementation Validation

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.