Academic Writing

figure-legend-writer

Writes complete, publication-grade figure legends that can stand on their own. Use when writing or revising figure legends for any scientific figure.

88100Total Score

Core Capability

94 / 100

Functional Suitability

12 / 12

Reliability

11 / 12

Performance & Context

6 / 8

Agent Usability

16 / 16

Human Usability

8 / 8

Security

12 / 12

Maintainability

11 / 12

Agent-Specific

18 / 20

Medical Task

15 / 15 Passed

90Bar chart: 3 groups (control, treatment A, treatment B), n=6/group, one-way ANOVA + Tukey, SEM error bars

5/5

853-panel figure: (A) western blot, (B) bar graph quantification, (C) microscopy — loading control not specified

5/5

76Minimal input: 'Write a legend for my figure' — no figure type or details

5/5

Veto GatesRequired pass for any deployment consideration

Skill Veto✓ All 4 gates passed

✓

Operational Stability

System remains stable across varied inputs and edge cases

PASS

✓

Structural Consistency

Output structure conforms to expected skill contract format

PASS

✓

Result Determinism

Equivalent inputs produce semantically equivalent outputs

PASS

✓

System Security

No prompt injection, data leakage, or unsafe tool use detected

PASS

Research Veto✅ PASS — Applicable

Dimension	Result	Detail
Scientific Integrity	PASS	No fabricated sample sizes, p-values, or statistical tests detected. Placeholder convention explicitly used for missing values.
Practice Boundaries	PASS	No clinical interpretations produced. Skill scope is figure legend writing only.
Methodological Ground	PASS	No methodological fallacies. Statistical notation conventions correctly applied.
Code Usability	N/A	No code generated; Mode A text-output skill.

Core Capability94 / 100 — 8 Categories

Functional Suitability

Full marks. 11 figure types covered with type-specific additional elements; 8 template variants in legend_templates.md; journal-specific requirements for Nature, Cell, and Science journals included.

12 / 12

100%

Reliability

Placeholder convention prevents fabrication; Step 3 quality checklist with 7 verifiable criteria; no explicit handling for supplementary figure legends vs. main figure legends.

11 / 12

92%

Performance & Context

Figure-type-specific elements table adds necessary but non-trivial token overhead; overall 3-step workflow is lean for the complexity covered.

6 / 8

75%

Agent Usability

Full marks. Type-specific elements table, standardized placeholder convention, journal-specific requirements, and per-panel multi-panel format all well-designed.

16 / 16

100%

Human Usability

Full marks. Description lists specific figure type names as triggers; immediately recognizable by any researcher needing to write a legend.

8 / 8

100%

Security

Full marks. Hard rules prohibit fabricated sample sizes, p-values, and statistical tests; placeholder convention enforces this at the output level.

12 / 12

100%

Maintainability

Both reference files (academic_style_guide.md, legend_templates.md) present and well-structured; journal-specific word limits and requirements will need periodic updates; no version tracking or 'last verified' date in reference files.

11 / 12

92%

Agent-Specific

Trigger precision via specific figure type names is excellent; placeholder composability enables downstream verification workflow; progressive disclosure could be more explicit for asking figure type before generating type-specific elements.

18 / 20

90%

Core Capability Total94 / 100

Medical TaskExecution Average: 83.7 / 100 — Assertions: 15/15 Passed

Canonical

Bar chart: 3 groups (control, treatment A, treatment B), n=6/group, one-way ANOVA + Tukey, SEM error bars

5/5 ✓

Variant A

3-panel figure: (A) western blot, (B) bar graph quantification, (C) microscopy — loading control not specified

5/5 ✓

Edge

Minimal input: 'Write a legend for my figure' — no figure type or details

5/5 ✓

Canonical✅ Pass

Bar chart: 3 groups (control, treatment A, treatment B), n=6/group, one-way ANOVA + Tukey, SEM error bars

5/5 assertions passed. Complete legend using bar chart template; all required universal and type-specific elements present.

Basic 38/40|Specialized 52/60|Total 90/100

✅A1Legend includes figure number, brief title, and what each bar represents

✅A2Sample size (n=6 per group) is explicitly stated

✅A3Statistical test (one-way ANOVA + Tukey) and significance thresholds are stated

✅A4Error bar type (SEM) is explicitly defined in the legend

✅A5Legend does not fabricate statistical results not provided by the user

Pass rate: 5 / 5

Variant A✅ Pass

3-panel figure: (A) western blot, (B) bar graph quantification, (C) microscopy — loading control not specified

5/5 assertions passed. Per-panel structure applied; loading control placeholder correctly inserted.

Basic 34/40|Specialized 51/60|Total 85/100

✅A1Output addresses each panel separately using (A), (B), (C) structure

✅A2Western blot panel legend includes loading control placeholder

✅A3Microscopy panel legend includes scale bar placeholder

✅A4Step 4 output notes all placeholders that need to be filled before submission

✅A5Legend does not fabricate the loading control protein or scale bar value

Pass rate: 5 / 5

Edge✅ Pass

Minimal input: 'Write a legend for my figure' — no figure type or details

5/5 assertions passed. Figure type and critical details requested before proceeding; no legend produced.

Basic 30/40|Specialized 46/60|Total 76/100

✅A1Skill does not produce a legend without knowing the figure type

✅A2Output asks for figure type, sample size, statistical test, and error bar type

✅A3Output does not invent a figure description or placeholder legend

✅A4Output lists the universal elements that will be needed regardless of figure type

✅A5Output communicates that type-specific elements will be applied once figure type is known

Pass rate: 5 / 5

Medical Task Total83.7 / 100

Key Strengths

11-figure-type coverage with type-specific additional elements is the most comprehensive figure legend framework in this skill family
Standardized placeholder convention ([n = X per group], [AUTHOR: specify error bar type]) prevents fabrication while keeping the output usable
Journal-specific requirements (Nature, Cell Press, Science) in academic_style_guide.md enable direct journal-targeted output
8 legend templates in legend_templates.md provide copy-ready starting points for the most common figure types