explain-paper-html
Lesson builder for academic papers. Use when Codex needs a deep teaching-oriented paper/PDF explanation, a self-contained interactive HTML tutorial, separate or connected explanations for multiple papers, or a reader-facing paper explanation requested by another skill. Route a 1–4-page factual lookup to lightweight PDF reading and a format-only PDF-to-Markdown request to a converter.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 3
- Forks
- 0
01 前言 Front Matter
- name
- explain-paper-html
- description
- Lesson builder for academic papers. Use when Codex needs a deep teaching-oriented paper/PDF explanation, a self-contained interactive HTML tutorial, separate or connected explanations for multiple papers, or a reader-facing paper explanation requested by another skill. Route a 1–4-page factual lookup to lightweight PDF reading and a format-only PDF-to-Markdown request to a converter.
02 资源文件 5
- 参考文档agents/openai.yaml259 B
- 参考文档references/explanation-contract.md4.5 KB
- 参考文档references/html-verification.md3.1 KB
- 参考文档references/paper-analysis.md3.0 KB
- 脚本scripts/verify_explanation.py17.1 KB
03 SKILL.md 正文
Explain Paper HTML
Build a lesson whose teaching claims stay traceable to the paper and whose final HTML has been exercised in a browser.
1. Resolve the contract
Resolve the input paper and accessible form, requested language, relationship among papers, reference page or visual style, output path, and need for related-work research. Honor an explicit output path. Otherwise use resources/YYYY-MM-DD-explanation-<paper-slug>.html when the project has resources/, or /tmp/YYYY-MM-DD-explanation-<paper-slug>.html. Ask only when an unresolved choice would materially change the result.
Completion criterion: The input, language, paper relationship, reference style, research scope, and output path are recorded.
2. Build the evidence ledger
Read references/paper-analysis.md before analyzing the paper. For a multi-section PDF, invoke or follow pdf-explore and persist the whole parse. Inspect rendered pages or figure crops where the text layer cannot establish visual evidence. Record a page or section source and evidence type for every material claim, equation, and number intended for the lesson.
Completion criterion: Every planned technical claim, equation, experimental number, and visually derived conclusion maps to paper evidence and an evidence type.
3. Extract the design grammar
Read references/explanation-contract.md before analyzing a reference page or designing the default page. When reference HTML is provided, inventory its information architecture, visual language, components, diagram conventions, interactions, responsive behavior, and pedagogical rhythm. Separate reusable style decisions from paper-specific content.
Completion criterion: A content outline and visual-component inventory identify the source of every major design decision.
4. Design the lesson
Reorder the evidence into the learning path background → intuition → mechanism → evidence → boundaries. Include positioning and a one-sentence conclusion, prerequisites, concrete toy values, the method/proof/system walkthrough, evidence-strength assessment, assumptions and likely misreadings, and five medium-difficulty multiple-choice questions. For a series, state what the previous paper established, what this paper adds, and which boundaries remain distinct.
Completion criterion: The outline lets a reader construct the correct mental model without first reading the full paper, and all five questions test that model.
5. Implement the explanation
Apply the loaded HTML and multi-paper contracts. Produce one responsive long-page file per paper with inline CSS, inline JavaScript, a table of contents, semantic evidence notes, HTML or inline-SVG diagrams, preserved-whitespace <pre> blocks, useful interactions, and explanatory quiz feedback. Use the requested language while retaining English technical terms where useful.
Completion criterion: Each output is a self-contained offline file with working navigation, five feedback-bearing questions, and previous/next links where the task is a series.
6. Verify and hand off
Read references/html-verification.md, run scripts/verify_explanation.py on every output, validate inline JavaScript, and exercise the browser matrix there. Recheck chart labels and numbers against the evidence ledger. Keep temporary artifacts outside the deliverable directory and remove them after inspection. Report every unrun check as unrun and set the run status to incomplete until all required checks pass.
Completion criterion: Static validation, JavaScript checks, desktop/mobile/tail rendering, quiz-state checks, extra-interaction checks, visual inspection, and evidence rechecks all pass for every file.
04 同仓库更多 Skills
workflow-skill-creator
Distills a completed user workflow or interaction into a reusable agent skill. Use when the user asks to turn their workflow, interaction, or multi-step process into a skill, or when they say "make this a skill", "create a skill from what we just did", "package this workflow" or similar. Do not use for creating skills from scratch without an existing workflow (use a generic skill-creator for that).
uv
Checks whether the uv Python package manager is installed and installs it if missing. Ensures uv is on PATH. Use when another skill requires uv as a prerequisite.
uniprot-database
Access protein metadata, function, taxonomy, and sequences across UniProtKB, UniParc, and UniRef. Use when searching for proteins, mapping identifiers, or retrieving functional annotations and publications. Don't use for sequence alignment, protein folding, or sequence similarity search (use specialized skills for those tasks).
unibind-database
Queries the UniBind database for experimentally validated transcription factor (TF) binding sites. Use when retrieving direct TF-DNA interaction datasets, downloading binding site coordinates (BED/FASTA) for local analysis, or listing available datasets by species, cell line, or TF name. Don't use to query specific intervals, locations, genes, motif models or expression data.
ucsc-conservation-and-tfbs
Fetch Evolutionary Conservation scores (phyloP, phastCons) and Transcription Factor Binding Sites (TFBS) from the UCSC Genome Browser. Use when analyzing whether genomic variants or regions are evolutionarily conserved, functionally important, or bounded by TF regulators across major projects (ENCODE, JASPAR, ReMap).
string-database
Query the STRING database for protein-protein interactions (PPIs), functional enrichment, and homology. Use when the user asks about interactions between specific proteins, interaction evidence, confidence scores, protein interaction partners, or pathway enrichments.