CtrlK
BlogDocsLog inGet started
Tessl Logo

scientific-writing

Core skill for the deep research and writing tool. Write scientific manuscripts in full paragraphs (never bullet points). Use two-stage process with (1) section outlines with key points using research-lookup then (2) convert to flowing prose. IMRAD structure, citations (APA/AMA/Vancouver), figures/tables, reporting guidelines (CONSORT/STROBE/PRISMA), for research papers and journal submissions.

69

Quality

86%

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

SKILL.md
Quality
Evals
Security

Quality

Content

72%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with strong reference structure, but it is verbose and its workflows lack explicit validation feedback loops. Tightening the redundant prose-vs-bullets messaging, the over-claiming visual-generation section, and Section 10's general-knowledge terminology would materially improve it.

Suggestions

State the 'write in full paragraphs, never bullet points' rule once in a prominent place and remove the repeated restatements across Overview, Section 7, the Workflow, and the Common Mistakes list to recover tokens.

Replace the 'MANDATORY/EXTENSIVELY/CRITICAL' visual-generation emphasis and the inflated minimum-figure table (e.g., Market Research: 20–30 figures) with a single concise, calibrated figure-generation guidance block.

Add explicit validation checkpoints to the manuscript workflow — e.g., 'compile the .tex and fix errors before proceeding' and 'verify every citation resolves to a real source' — to create verify-then-proceed feedback loops.

Move Section 10's field-by-field terminology into a reference file (e.g., references/field_terminology.md) and keep only the general principles inline, since much of that content is knowledge Claude already has.

DimensionReasoningScore

Conciseness

Mostly useful but padded: the 'always write in full paragraphs, never bullet points' rule is repeated 5+ times, the visual-enhancement section over-emphasizes with MANDATORY/EXTENSIVELY/CRITICAL and an over-claiming minimum-figure table, and Section 10 (~90 lines) explains field-specific terminology Claude largely already knows.

2 / 3

Actionability

Provides concrete, executable guidance throughout — bash commands for generate_schematic.py/generate_image.py and complete, copy-paste-ready LaTeX for box environments, tables, scientific-notation commands, and the getting-started template.

3 / 3

Workflow Clarity

The Stage 1–4 manuscript workflow is clearly sequenced and includes a Stage 3 revision checklist, but it lacks explicit verify-then-proceed gating and error-recovery feedback loops (e.g., compile-check-fix for LaTeX, citation-resolution verification).

2 / 3

Progressive Disclosure

Major topics are split into real, well-signaled one-level-deep reference files (imrad_structure.md, citation_styles.md, figures_tables.md, reporting_guidelines.md, writing_principles.md, professional_report_formatting.md) and assets, with a dedicated References/Assets listing for easy navigation.

3 / 3

Total

10

/

12

Passed

Description

100%

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 strong, specific description that names concrete actions, uses natural domain trigger terms, and explicitly signals when to use it. The only soft spot is the vague branding opener 'Core skill for the deep research and writing tool', which adds little but does not undermine the dimension scores.

DimensionReasoningScore

Specificity

Lists multiple concrete actions such as 'Write scientific manuscripts', a 'two-stage process with (1) section outlines... (2) convert to flowing prose', and 'IMRAD structure, citations (APA/AMA/Vancouver), figures/tables, reporting guidelines (CONSORT/STROBE/PRISMA)' rather than vague language.

3 / 3

Completeness

Explicitly states what the skill does and provides equivalent explicit trigger guidance via 'for research papers and journal submissions', so it is not capped at 2 and clearly answers both what and when.

3 / 3

Trigger Term Quality

Covers natural terms a researcher would actually say — 'scientific manuscripts', 'research papers', 'journal submissions', 'IMRAD', 'citations', and the reporting-guideline acronyms CONSORT/STROBE/PRISMA.

3 / 3

Distinctiveness Conflict Risk

The scientific-manuscript/journal-submission niche with IMRAD and CONSORT/STROBE/PRISMA triggers is clearly distinguishable from general writing skills and unlikely to trigger the wrong skill.

3 / 3

Total

12

/

12

Passed

Validation

81%

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

Validation13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (718 lines); consider splitting into references/ and linking

Warning

metadata_field

'metadata' should map string keys to string values

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

13

/

16

Passed

Repository
K-Dense-AI/claude-scientific-writer
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.