data-transform

Transform, clean, reshape, and preprocess data using pandas and numpy. Works with ANY LLM provider (GPT, Gemini, Claude, etc.).

Quality

58%

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/Data Analysis/data-transform/SKILL.md

Quality

Content

50%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 comprehensive, mostly executable pandas/numpy cookbook, but it is noticeably verbose with basic operations Claude already knows and lacks validation checkpoints in its batch/destructive workflows (capping workflow clarity at 3). With no bundle files, all detail is inlined rather than progressively disclosed.

Suggestions

Trim or move basic pandas operations Claude already knows into a separate reference file, keeping only non-obvious or domain-specific patterns (AnnData, DEG processing, CPM normalization) inline.

Add explicit validation checkpoints (e.g., assert/inspect outputs after destructive drops and after batch file combination) with fix-and-retry guidance to raise workflow clarity above 3.

Split the Advanced Features and Common Use Cases sections into referenced bundle files (e.g., ADVANCED.md, USE_CASES.md) with clearly signaled links for progressive disclosure.

Dimension	Reasoning	Score
Conciseness	The ~590-line body extensively explains basic pandas operations Claude already knows (isnull/drop_duplicates, melt/pivot, merge, groupby) and is padded with print statements and emoji checkmarks ("✅ Data cleaned"), matching 'Noticeably verbose; several unnecessary explanations or padded sections'. Not a 1 because the content is accurate and not purely conceptual padding; not a 3 because the volume of redundant basic instruction is clearly excessive.	2 / 5
Actionability	Provides extensive concrete, mostly copy-paste-ready code across cleaning, scaling, reshaping, merging, and feature engineering, matching 'Mostly executable guidance; concrete code or commands with minor gaps'. Not a 5 because several snippets assume undefined variables (df1/df2, adata, deg_df); not a 3 because the bulk is real executable code rather than pseudocode.	4 / 5
Workflow Clarity	Steps 1-6 are sequenced under 'How to Use', but validation is only a vague Best-Practices bullet ('Check data after each major transformation') with no checkpoints, and the batch/destructive operations lack validate-fix-retry loops, so per the batch/destructive cap it cannot exceed 3. Not a 4 because explicit validation checkpoints and error-recovery loops are absent.	3 / 5
Progressive Disclosure	Sections are organized (Overview, When to Use, How to Use, Advanced, Common Use Cases) but no bundle files exist and all API-reference-style content is inlined in one file with only external URLs in References, matching 'Some structure but could be better organized; content that should be separate is inline'. Not a 4 because there are no signaled one-level-deep reference files to offload detail; not a 2 because section structure is present rather than minimal.	3 / 5
	Total	12 / 20 Passed

Description

66%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 states concrete capabilities and named libraries clearly, but omits any 'Use when...' trigger guidance, which caps completeness at 3. The LLM-provider compatibility framing is irrelevant to triggering and dilutes trigger-term quality.

Suggestions

Add an explicit 'Use when...' clause naming natural trigger phrases (e.g., 'Use when cleaning, normalizing, reshaping, or preprocessing tabular data with pandas/numpy').

Remove the 'Works with ANY LLM provider (GPT, Gemini, Claude, etc.)' framing — it is not a trigger term and does not help distinguish the skill's purpose.

Add concrete synonym/extension terms (normalize, scale, merge, ETL, CSV/Excel) to improve trigger-term coverage toward a 5.

Dimension	Reasoning	Score
Specificity	Quotes "Transform, clean, reshape, and preprocess data using pandas and numpy" — four named actions plus named libraries, matching the anchor 'Lists several specific actions; minor gaps in coverage'. It is not a 3 (only 1-2 actions) nor a 5 (the actions like 'clean'/'preprocess' are broad rather than comprehensive concrete operations).	4 / 5
Completeness	It gives a clear 'what' but no 'when' — the second sentence ("Works with ANY LLM provider") addresses provider compatibility, not trigger guidance, so per the guideline a missing 'Use when...' clause caps completeness at 3. Not a 4 because 'when' is entirely absent, not merely weak.	3 / 5
Trigger Term Quality	Includes natural analyst terms ("clean", "transform", "reshape", "preprocess") users would say, matching 'Good keyword coverage; a few natural terms missing'. Not a 5 because it omits common synonyms/extensions (e.g., 'normalize', 'ETL', file types) and spends tokens on provider names rather than triggers; not a 3 because coverage is genuinely good.	4 / 5
Distinctiveness Conflict Risk	The pandas/numpy data-preprocessing niche is mostly distinct with only minor overlap risk against general data skills, matching 'Mostly distinct; minor overlap risk'. Not a 5 due to some overlap with broad data-cleaning skills and an off-topic provider framing; not a 3 because it is not broadly generic.	4 / 5
	Total	15 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (589 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	14 / 16 Passed

Repository: aipoch/medical-research-skills
Path: scientific-skills/Data Analysis/data-transform/SKILL.md
Commit: f5ef65b

Reviewed: ~2 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.