/journal-club-ppt

name: journal-club-ppt
description: "Use this skill whenever the user provides a scientific paper PDF and asks for a group-meeting literature report, journal-club slides, 文献汇报PPT, 组会PPT, paper presentation, article walkthrough, or to explain a paper with PowerPoint. The skill first reconstructs the paper's scientific logic, then builds author/background sections, chooses an evidence-driven slide outline, crops only main-text figure panels from the PDF, and creates an academic PPT with 10–30 slides including title and conclusion/discussion. Always use this skill for '把这篇论文做成PPT', '文献汇报', 'journal club', or '组会汇报', even if the user only says they need slides."

Journal Club PPT — scientific paper → academic group-meeting slides

This skill turns a scientific paper PDF into a claim-driven PPT for lab meeting, journal club, or 文献汇报. The deck must explain the paper's logic, not merely summarize sections or paste screenshots.

Default language: Chinese, with key technical terms preserved in English when helpful. Use another language only if the user asks.

Hard constraints

1. **Read the argument before designing slides.** Do not create a slide outline until you can state the paper's central question, knowledge gap, hypothesis, experimental/computational strategy, evidence chain, and final claim. 2. **Include author introduction.** At minimum: first author(s), corresponding/senior author(s), affiliations, research direction inferred from the paper, and why this team is relevant to the work. Use external sources only for text facts when available and appropriate; never import external images. 3. **Include background introduction.** Explain the field context as: broad problem → unresolved gap → why this paper matters → what question the paper asks. 4. **Confirm a PPT outline before filling slides unless the user asked for a fully automatic final deck.** The outline must be evidence-driven and include planned figure panels. 5. **Use only figures from the PDF's main text/body.** Do not use web images, stock images, AI-generated images, author photos, journal logos, graphical abstracts, supplementary figures, Extended Data, or figures from other papers unless the user explicitly overrides this constraint. 6. **Select subfigures by logic.** Crop and use only panels that support the narrative. Do not paste whole pages or dump every panel from every figure. 7. **End with evaluation.** The deck must discuss strengths, limitations, unanswered questions, and possible follow-up experiments/analyses. 8. **Slide count is a hard quality gate.** The final PPT must contain **at least 10 and at most 30 slides**, including title and conclusion/discussion slides. Use 10–12 slides for short papers, 14–18 for standard research articles, 19–24 for complex multi-omics or method-heavy papers, and 25–30 only when the paper genuinely needs it.

Wisp-specific workflow

Use Wisp's strengths rather than treating the PDF as static screenshots:

• Use PDF text extraction / `pdf-explore`-style helpers when available to read abstract, introduction, results, methods, discussion, captions, author information, and references.
• Use vision on rendered PDF pages to inspect figures, layouts, panel labels, legends, axes, scale bars, and multi-panel structure.
• Use Python to render PDF pages at high DPI, crop selected figure panels, name crops consistently, and build a figure ledger.
• Save the paper outline, figure inventory, crop paths, and slide plan in project files. Python work can run as standalone scripts or in Wisp's persistent runtime; choose according to the workflow and whether retaining loaded PDF/image objects or helpers benefits subsequent steps.
• Use MCP/web/PubMed/ORCID/lab pages only for textual author/background verification when helpful. External visual assets remain forbidden.

Phase 1 — Paper logic reconstruction

Before touching the PPT, produce a compact internal `paper_logic` brief:

• **Bibliographic identity:** title, journal/preprint server, year, DOI if present, article type, field.
• **Central problem:** what biological/clinical/computational problem is being addressed?
• **Knowledge gap:** what prior uncertainty or technical bottleneck motivates the work?
• **Main hypothesis or central claim:** one sentence.
• **System and data:** organism/cell type/cohort/dataset/model, sample scale, intervention or comparison.
• **Experimental/computational design:** key assays, models, algorithms, controls, statistics.
• **Evidence chain:** ordered claims from Result 1 → Result 2 → Result N, each mapped to figure panels.
• **Final conclusion:** what becomes more believable after the paper?
• **Scope boundary:** what the paper does *not* prove.

Use the whole paper, especially results opening/closing paragraphs and figure captions. Do not rely only on the abstract.

Phase 2 — Author and background preparation

Author slide

Create one author slide unless the user asks for a longer introduction. Prefer facts from the PDF:

• First author(s): role inferred from author order and contribution statement if present.
• Corresponding/senior author(s): affiliation, lab/institute, likely research area from the article and author contribution statement.
• Collaboration structure: single-lab, multi-center, consortium, clinical cohort, computational/experimental partnership.
• Why the author team matters for this paper: access to special cohort, platform, model system, algorithm, or prior expertise.

Avoid unsupported reputation claims such as “国际顶尖专家” unless sourced. If using external textual sources, record source links in speaker notes or a references slide. Do not use author portraits or lab logos unless they are in the PDF body figures and are scientifically relevant, which is rare.

Background slides

The background must teach enough context for a lab audience to follow the results:

1. Define the biological/technical problem. 2. Explain the state of the field before this paper. 3. Identify the