scientific-slides
Build slide decks and presentations for research talks. Use this for making PowerPoint slides, conference presentations, seminar talks, research presentations, thesis defense slides, or any scientific talk. Provides slide structure, design templates, timing guidance, and visual validation. Works with PowerPoint and LaTeX Beamer.
75
67%
Does it follow best practices?
Impact
89%
4.45xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Optimize this skill with Tessl
npx tessl skill review --optimize ./scientific-skills/scientific-slides/SKILL.mdQuality
Discovery
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.
This is a strong skill description that clearly defines its scope (research and scientific presentations), lists concrete capabilities (structure, templates, timing, validation), and provides extensive trigger terms covering many natural user phrasings. It effectively answers both what the skill does and when to use it, with a well-defined niche that minimizes conflict risk with other skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'Build slide decks and presentations', 'slide structure, design templates, timing guidance, and visual validation'. Also specifies concrete tools: 'PowerPoint and LaTeX Beamer'. | 3 / 3 |
Completeness | Clearly answers both what ('Build slide decks and presentations... Provides slide structure, design templates, timing guidance, and visual validation') and when ('Use this for making PowerPoint slides, conference presentations, seminar talks, research presentations, thesis defense slides, or any scientific talk') with explicit trigger guidance. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'slide decks', 'presentations', 'PowerPoint slides', 'conference presentations', 'seminar talks', 'research presentations', 'thesis defense slides', 'scientific talk', 'LaTeX Beamer'. These are highly natural phrases a researcher would use. | 3 / 3 |
Distinctiveness Conflict Risk | Clearly scoped to research/academic presentations specifically, with distinct triggers like 'thesis defense', 'conference presentations', 'seminar talks', 'LaTeX Beamer'. This niche focus on scientific/research talks distinguishes it well from a generic presentation skill. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is comprehensive in scope but severely undermined by extreme verbosity and repetition. The same core advice (visual-first design, minimal text, use research-lookup, modern color palettes) is repeated 4-5 times across different sections. While the Nano Banana Pro script examples are concrete and actionable, much of the surrounding content explains concepts Claude already knows (what bar charts are for, how to practice a talk, what sans-serif fonts are). The document would be dramatically more effective at 30-40% of its current length with design principles and talk-type guidance moved entirely to the referenced files.
Suggestions
Cut content by 60-70%: Remove all repeated advice (visual-first, minimal text, research-lookup guidance appears 4-5 times), eliminate explanations of concepts Claude knows (what conference talks are, what bar charts show, basic typography), and trust the referenced files to carry detailed guidance.
Consolidate workflows: The document has overlapping workflow descriptions in 'Slide Generation with Nano Banana Pro', 'Implementation Options', 'Workflow for Presentation Development', and 'Quick Start Guide'. Pick one canonical workflow per output format and remove duplicates.
Move design principles, talk-type guidance, timing details, and pitfalls entirely to the referenced files (references/slide_design_principles.md, references/talk_types_guide.md, etc.) rather than duplicating them inline AND referencing them externally.
Add explicit validation gates: Integrate validate_presentation.py output into the workflow as a decision point (e.g., 'If validation reports issues, fix before proceeding') rather than listing it as a separate tool.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Extremely verbose at ~600+ lines. Extensively explains concepts Claude already knows (what a conference talk is, what bar charts are for, what PDF means, basic design principles like 'use sans-serif fonts'). Massive repetition throughout—the same advice about '3-4 bullets, 4-6 words each', 'visual-first', 'use research-lookup for 8-15 papers', and color palette guidance is repeated 4-5 times across different sections. The anti-pattern/pitfalls section largely restates the design principles section. | 1 / 3 |
Actionability | Provides concrete bash commands for the generate_slide_image.py and slides_to_pdf.py scripts with executable examples, which is good. However, much of the content is advisory rather than executable (design principles, timing advice, practice schedules). The scripts referenced (validate_presentation.py, generate_schematic.py) are not verified to exist since no bundle files are provided. The PPTX workflow defers to another skill without concrete steps. | 2 / 3 |
Workflow Clarity | The 6-stage workflow is clearly sequenced and the PDF generation workflow has explicit steps. However, the visual validation workflow (Stage 4) lacks a true feedback loop with explicit stopping criteria tied to automated checks—the validate_presentation.py script is mentioned but its output isn't integrated into a decision gate. The workflow is also buried within an enormous document, making it hard to follow. Multiple overlapping workflow descriptions (Quick Start, Stage 2, Implementation Options) create confusion about which to follow. | 2 / 3 |
Progressive Disclosure | References to external files are well-signaled (references/*.md, assets/*.tex, etc.) and organized by category. However, since no bundle files are provided, we can't verify these exist. The main problem is that the SKILL.md itself is monolithic—it contains enormous amounts of inline content that should be in the referenced files (e.g., detailed design principles, talk-type guidance, timing details are all inline AND referenced to external files). The document tries to be both overview and comprehensive guide simultaneously. | 2 / 3 |
Total | 7 / 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.
Validation — 9 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (1155 lines); consider splitting into references/ and linking | Warning |
metadata_version | 'metadata.version' is missing | Warning |
Total | 9 / 11 Passed | |
- Repository
- K-Dense-AI/claude-scientific-skills
- Commit
cbcae7b
- 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.