CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

exploratory-data-analysis

Perform comprehensive exploratory data analysis on scientific data files across 200+ file formats. This skill should be used when analyzing any scientific data file to understand its structure, content, quality, and characteristics. Automatically detects file type and generate...

Quality

71%

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/exploratory-data-analysis/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

67%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 delivers a genuinely usable EDA workflow: clear step sequencing, a runnable analyzer script, well-signaled one-level-deep references that all exist, per-datatype analysis checklists, and solid troubleshooting for missing libraries, unknown formats, and large files. Its weaknesses are redundancy and example quality — three sections repeat content (format-lookup instructions, script usage, the reference-file listing), and the worked examples are comment skeletons rather than executable code. Trimming the duplicated sections and tightening the examples would raise both conciseness and actionability without losing information.

Suggestions

Remove or merge the duplicated sections: 'Detailed Format References' repeats Workflow Step 2, 'Script Usage' repeats Step 3 Option A, and 'Resources' re-lists the reference files a third time — keep one of each.

Replace the comment-skeleton Examples with executable code that actually computes and prints the advertised metrics (read count, length distribution, GC content), so they are copy-paste runnable.

Drop the per-category inline extension lists (e.g., the ~10 extensions listed per category) since the reference files already enumerate them; keep only the category-to-file mapping.

DimensionReasoningScore

Conciseness

The bulk is operational (workflow, per-datatype analysis checklists, troubleshooting), but there is real padding: 'Detailed Format References' (lines 200-232) repeats Workflow Step 2, 'Script Usage' (382-400) repeats Step 3 Option A, the 'Resources' section re-lists the reference files a third time, and each category section inlines ~10 example extensions that the reference files already cover. This sits between anchors 2 and 3 — mostly efficient with several sections that could be cut or tightened — so the even score 3 reflects that it is at the efficient side of the boundary; it is above anchor 2 because the core guidance is concrete and does not explain concepts Claude already knows, and below anchor 4 because the duplicated sections are unnecessary tokens.

3 / 5

Actionability

There is an executable command ('python scripts/eda_analyzer.py <filepath> [output.md]'), a concrete regex for searching large reference files, specific per-category library lists (biopython, pysam, tifffile, nd2reader, ...), real import statements in the examples (SeqIO.parse, pd.read_csv, ND2Reader), and a report template file. This matches anchor 4 (mostly executable guidance with minor gaps); below anchor 5 because the three worked Examples are largely comment skeletons ('# Calculate: read count, length distribution...') rather than copy-paste runnable code, above anchor 3 because the concrete commands and library specifics go beyond pseudocode.

4 / 5

Workflow Clarity

A clear five-step sequence (detect extension → load reference info → analyze via script or custom code → generate report → save) is presented with an inline worked example, plus troubleshooting paths for missing libraries, unknown file types, and large files. This matches anchor 4 (clear sequence, most checkpoints present, minor validation gaps); below anchor 5 because there is no explicit validate-then-retry loop on the generated report (e.g., re-checking that detected dimensions match stated metadata is mentioned only as a best-practice bullet), above anchor 3 because the sequence is well-defined and error cases are addressed — appropriate for a read-only analysis and report-writing operation that is not destructive or batch-risky.

4 / 5

Progressive Disclosure

The body maps six categories to six real, one-level-deep reference files (all exist and use '### .ext' section headings), signals the script and report template by path, and explicitly instructs how to search the large reference files without loading them whole. This matches anchor 4 (good structure, references mostly clear, minor organization gaps); below anchor 5 because the body inlines extension lists that duplicate the reference files and repeats the reference-file listing three times (category sections, 'Detailed Format References', 'Resources'), above anchor 3 because navigation to the bundle is consistently well-signaled and easy to follow.

4 / 5

Total

15

/

20

Passed

Description

62%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 correctly uses third person, states a clear purpose, and includes an explicit 'use when' clause, so it is functional as a trigger. Its main defects are a truncated final sentence that cuts the capability list mid-action ('Automatically detects file type and generate...') and a lack of natural trigger synonyms or file extensions, which weakens both keyword coverage and distinctiveness. Completing the sentence and adding user-facing terms (e.g., 'explore', 'EDA', '.csv', '.fastq') would lift it substantially.

Suggestions

Complete the truncated final sentence — e.g., 'Automatically detects file type and generates a comprehensive markdown report with statistics, quality assessment, and downstream analysis recommendations.'

Add natural trigger keywords and synonyms users would say, such as 'explore', 'EDA', 'summarize', or 'profile a dataset', so the skill surfaces for those phrasings.

Mention a few concrete file extensions (e.g., .csv, .xlsx, .fastq, .mzML) to sharpen trigger matching and distinguish the skill from generic data-analysis skills.

DimensionReasoningScore

Specificity

The description names the domain ('scientific data files across 200+ file formats') and a couple of concrete action facets ('understand its structure, content, quality, and characteristics', 'Automatically detects file type'), but the final action is cut off mid-sentence ('Automatically detects file type and generate...') and 'comprehensive' is filler rather than a capability. This matches anchor 3 (domain plus 1-2 concrete actions, not comprehensive); it is above anchor 2 because the actions are more concrete than 'Processes PDF files', and below anchor 4 because no list of several distinct specific operations survives the truncation.

3 / 5

Completeness

It has an explicit 'when' clause ('This skill should be used when analyzing any scientific data file to understand its structure, content, quality, and characteristics') and a stated 'what' ('Perform comprehensive exploratory data analysis... Automatically detects file type and generate...'), but the 'what' is truncated mid-action and the 'when' is generic ('any scientific data file'). This matches anchor 4 (both what and when present, when could be more explicit); below anchor 5 because the incomplete final sentence leaves the capability list unfinished, above anchor 3 because the trigger clause is explicit rather than weakly implied.

4 / 5

Trigger Term Quality

Relevant keywords are present — 'exploratory data analysis', 'analyzing', 'scientific data file' — but common natural variations users would actually say are missing: 'explore', 'EDA', 'summarize', 'profile', and file extensions like .csv, .fastq, or .mzML. This matches anchor 3 (some relevant keywords but missing common variations or synonyms); below anchor 4 because natural phrasings users say when handing over a data file are absent, above anchor 2 because the terms included are domain-appropriate rather than generic.

3 / 5

Distinctiveness Conflict Risk

The niche is fairly clear — EDA of scientific data files in 200+ (implicitly scientific) formats — which distinguishes it from general document or code skills, though 'analyzing data files' still overlaps with generic data-analysis skills and the description gives no extensions or scope boundaries that would sharpen the distinction. This matches anchor 4 (mostly distinct, minor overlap risk with closely related skills); below anchor 5 because the truncated, extension-free wording leaves the niche less sharply delimited than the anchor example, above anchor 3 because 'scientific data files' plus the EDA framing is more specific than 'Works with document files'.

4 / 5

Total

14

/

20

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.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

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/Data Analysis/exploratory-data-analysis/SKILL.md

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.