result-figure-consistencycheck

Checks consistency between paper result descriptions and figure legends (text-only) when the input is a PDF-to-Markdown full text containing page breaks (e.g., `## Page XX`) and legend text; outputs a Markdown consistency report and a UTF-8 CSV issue list.

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/Other/result-figure-consistencycheck/SKILL.md

Quality

Content

53%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 well-structured with a clear scope, I/O contract, and appropriate bundle references, but it leans on generic boilerplate, gives only high-level comparison guidance, and omits inline validation checkpoints for an output-writing workflow. An orphaned bundle CSV also hurts navigation.

Suggestions

Replace generic boilerplate (Output Contract, Failure Handling, User Checkpoints, Input Validation, Quick Validation) with skill-specific guidance, and de-duplicate the output-format details that appear in both Example Usage and Implementation Details.

Add concrete comparison heuristics — e.g. how to match a legend block to its Results paragraph by Figure number, what fields to diff (group names, time points, units, n-values) — so the core step is executable rather than conceptual.

Insert an inline validation/feedback checkpoint in the workflow (e.g. verify each recorded issue is traceable to a Page XX citation before writing outputs) to lift workflow_clarity above the destructive/batch cap of 3, and add a navigation pointer to assets/figure_caption_checklist.csv so no bundle file is orphaned.

Dimension	Reasoning	Score
Conciseness	Core sections (When to Use, Key Features, Example Usage, Implementation Details) are efficient and avoid over-explaining known concepts, but generic boilerplate (Required Inputs, Output Contract, Failure Handling, User Checkpoints, Input Validation, Quick Validation) plus redundancy between Example Usage and Implementation Details could be tightened.	3 / 5
Actionability	Concrete I/O contract is given (CSV columns, output file paths, template reference), but the core analytical step is only high-level — "Compare legend statements against Results statements and record discrepancies" — with no specific comparison methodology or matching heuristics to execute.	3 / 5
Workflow Clarity	A 4-step conceptual workflow is sequenced (Read → Identify → Compare → Write), but because the skill writes/overwrites output files, missing inline validation/feedback loops caps workflow_clarity at 3 per the rubric's destructive/batch cap.	3 / 5
Progressive Disclosure	Good structure with clearly signaled one-level-deep references to references/guide.md and assets/consistency_template.md; the one organization gap is that assets/figure_caption_checklist.csv is bundled but never referenced or navigated to from the body.	4 / 5
	Total	13 / 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 is specific and occupies a distinct niche, clearly stating what it does and what it outputs. Its main weakness is the missing explicit "Use when..." trigger clause and absent common synonyms (e.g. "caption", "discrepancy") that users would naturally say.

Suggestions

Add an explicit "Use when..." clause with natural user triggers, e.g. "Use when verifying that Results text matches figure captions/legends, or when checking for discrepancies between main-text results and figure legends."

Include common synonyms a user might say — "figure caption", "caption", "mismatch", "discrepancy" — alongside "figure legends" and "consistency".

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete actions — "Checks consistency between paper result descriptions and figure legends", "outputs a Markdown consistency report and a UTF-8 CSV issue list" — with only minor coverage gaps (panel-level granularity not surfaced).	4 / 5
Completeness	The "what" is clear, but the "when" is only an input precondition ("when the input is a PDF-to-Markdown full text containing page breaks") with no explicit "Use when..." trigger clause, capping completeness at 3 per the rubric guideline.	3 / 5
Trigger Term Quality	Has relevant niche keywords ("figure legends", "result descriptions", "consistency", "CSV") but misses common synonyms a user would naturally say, e.g. "caption"/"figure caption", "mismatch", "discrepancy".	3 / 5
Distinctiveness Conflict Risk	A clear niche — text-only consistency checking between paper Results text and figure legends from PDF-to-Markdown input — with distinct triggers and minimal risk of firing for an unrelated skill.	5 / 5
	Total	15 / 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/Other/result-figure-consistencycheck/SKILL.md
Commit: 63c61d3

Reviewed: 7 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.