跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
演示文稿

paperbanana

Use when user needs academic diagrams, methodology figures, statistical plots, or presentation slides from text descriptions or data files. Also use for evaluating generated figures against references.

PlutoLei · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add PlutoLei/paperbanana-skill --skill paperbanana

技能说明


name: paperbanana description: Use when user needs academic diagrams, methodology figures, statistical plots, or presentation slides from text descriptions or data files. Also use for evaluating generated figures against references. argument-hint: "[generate|plot|slide|slide-batch|evaluate|data|setup] [description or file path]" allowed-tools: Read, Write, Bash, Glob, Grep, AskUserQuestion

PaperBanana - Academic Illustration Generator

Multi-agent pipeline (Retriever → Planner → Stylist → Visualizer → Critic) for publication-quality academic diagrams, statistical plots, and presentation slides.

API key: Set provider keys in PaperBanana project's .env file. Timeout: 300000 (5 min) for all generation commands.


Commands

All commands run from project root: cd <paperbanana_dir> && python -m paperbanana.cli <cmd>

Command Selection Decision Tree

Route user requests to the right subcommand before looking up parameters:

User intentSignal wordsSubcommand
方法论/架构/流程图 from text or PDF"method figure", "架构图", "流程图", "methodology", "pipeline diagram", "论文配图"generate
Statistical plot from data file"plot", "curve", "bar chart", "scatter", "heatmap", has CSV/JSONplot
Single presentation slide"slide", "一张幻灯片", "封面图", single prompt fileslide
Batch slide generation"all slides", "批量生成", "N 张幻灯片", prompts/ directoryslide-batch
Compare generated vs human reference"evaluate", "对比", "与参考图对比"evaluate
Manage reference dataset"download dataset", "清缓存"data
First-time provider config"setup", "配置 API key"setup

Ambiguous input: If user provides just a description with no subcommand signal, default to generate (see Argument Parsing table for details).

Out-of-scope: Pure code generation (matplotlib/seaborn script) is NOT paperbanana's job — those go to matplotlib / scientific-visualization skills. Paperbanana is for AI-driven image generation + critique loops.

Note (upstream sync pending): Upstream paperbanana CLI also adds subcommands (plot-batch #123, sweep #118) not yet reflected in this table. See the llmsresearch/paperbanana CHANGELOG for the authoritative CLI surface.

generate — Methodology Diagrams

python -m paperbanana.cli generate --input '<file>' --caption '<caption>' --optimize --verbose

When user provides inline text (no file): write to temp file, use as --input.

ParameterDefaultDescription
--input / -i—Path to methodology text file or PDF (.pdf requires pip install 'paperbanana')
--caption / -c—Figure caption / communicative intent
--output / -oautoOutput image path
--vlm-providergeminiVLM provider: gemini, anthropic, openai, bedrock, openrouter, ollama, claude_code, litellm
--vlm-modelautoVLM model name
--image-providerautoImage gen provider: google_imagen, openai, bedrock, openrouter
--image-modelautoImage gen model name
--iterations / -n3Max critic rounds
--autooffLoop until critic is satisfied (safety cap via --max-iterations)
--max-iterations30Safety cap for --auto mode
--optimizeoffPreprocess inputs (parallel enrichment + caption sharpening)
--continueoffContinue from the latest run
--continue-run—Continue from a specific run ID
--feedback—User feedback for the critic when continuing a run
--aspect-ratio / -arautoTarget aspect ratio: 1:1, 2:3, 3:2, 3:4, 4:3, 9:16, 16:9, 21:9
--format / -fpngOutput format: png, jpeg, webp
--dry-runoffValidate inputs without making API calls
--exemplar-retrievaloffEnable external exemplar retrieval before planning
--seed—Random seed for reproducible generation
--verbose / -voffShow detailed agent progress and timing
--auto-download-dataoffAuto-download expanded reference set (~257MB) on first run
--venue—Academic venue style: neurips, icml, acl, ieee, custom
--pages—Page range for PDF input (e.g., 3-5)
--config—Path to config YAML file

Venue styles: --venue neurips applies NeurIPS-specific methodology and plot style guides from data/guidelines/. Each venue has distinct color palettes, layout conventions, and typography expectations.

PDF input: --input paper.pdf --pages 3-5 extracts text from the specified pages as source context.

Exemplar advanced flags: --exemplar-retrieval enables retrieval; see generate --help for additional config flags (--exemplar-endpoint, --exemplar-mode, --exemplar-top-k, --exemplar-timeout, --exemplar-retries).

plot — Statistical Plots

python -m paperbanana.cli plot --data '<data.csv>' --intent '<intent>' --optimize --verbose
ParameterDefaultDescription
--data / -d—Path to data file (CSV or JSON) [required]
--intent—Communicative intent for the plot [required]
--output / -oautoOutput image path
--vlm-providergeminiVLM provider
--iterations / -n3Refinement iterations
--format / -fpngOutput format
--aspect-ratio / -arautoTarget aspect ratio
--optimizeoffEnrich context and sharpen caption
--autooffLoop until critic satisfied
--verbose / -voffDetailed progress

slide — Presentation Slides

python -m paperbanana.cli slide --input '<prompt.md>' --resolution 4k
ParameterDefaultDescription
--input / -i—Path to slide prompt markdown
阅读完整来源 ↗