Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →cnfjlhj/paper2html
Use when turning an academic paper PDF/arXiv/OpenReview page/local LaTeX source into a single-file Chinese HTML deep-reading page, especially when the user wants a Cheat-Sheet style explanation, KaTeX formulas, figure/table evidence, validation, or a publishable research page.
Passed 16 of 19 test cases with the skill, and 3 of 19 without it.
and 23 more in the full report
14 fixed4 passed either way4 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
I need to initialize a local paper reading workspace for arXiv paper 2504.07952 (titled 'Agent System Architecture', GitHub https://github.com/org/repo). I want to set up the workspace output directory at /tmp/workspace, assign slug 'agent-arch', and set the publication target to 'ai/agent/arch'. What command should I execute to scaffold this project?
| Checked | Without skill | With skill |
|---|---|---|
| The generated command invokes scripts/bootstrap_paper2html.py. | Missed | Passed |
| The command uses the --arxiv flag set to 2504.07952. | Passed | Passed |
| The command includes --publish-to set to ai/agent/arch. | Missed | Passed |
The instructions your model reads when the skill loads.
Create one self-contained Chinese index.html that lets the reader understand the paper without reopening the PDF. Open with a compact, source-faithful story, then deliver a clear Cheat-Sheet page with evidence boundaries, KaTeX formulas, complete figure coverage, reviewer-level analysis, and no decorative complexity.
Every completed page ends its substantive analysis with all seven MIT Other Discussion Roles. Each role shows the original English prompt verbatim and a source-grounded Chinese response.
Use scripts first. Keep this skill as workflow and quality gates, not a long prompt.
cd /path/to/paper2html
# 1) Create workspace, notes, and starter index.html
python scripts/bootstrap_paper2html.py \
--title "Paper Title" \
--slug paper-slug \
--out /path/to/output/paper-slug-html-YYYYMMDD \
--arxiv 2504.07952 \
--github https://github.com/org/repo \
--publish-to topic/series/paper-slug
# 2) Inventory TeX/source when available
python scripts/tex_inventory.py \
/path/to/output/source/unpacked \
--out /path/to/output/notes/tex-inventory.md \
--json /path/to/output/notes/tex-inventory.json
# 3) Optimize figures and create data URIs for single-file HTML
python scripts/optimize_paper_images.py \
paper_figures/*.pdf paper_figures/*.png \
--out-dir /path/to/output/assets/optimized \
--width 1600 \
--data-uri-json /path/to/output/assets/data-uris.json
# 4) Optional: ask Gemini for a visual draft on a copy
scripts/gemini_frontend_pass.sh \
/path/to/output/index.html \
test cases passed, vs 3 / 19 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.