visual-diagrams
Skill1 starsChanged 14 days agoMITView on GitHub
把内容做成一张图:信息卡长图、示意图/流程图、技术架构图。按输出类型路由三条路径——① 信息卡与长图海报走 HTML/CSS + Playwright 出 PNG(深色杂志风设计规范);② 示意图/流程图/电路图手写 SVG,Chrome 渲染后上传思源;③ 技术架构图调用外部 fireworks 风格库(12 种现成风格)。含命名、导出、校验的通用约定。Triggers: 做一张图, 信息卡, 长图, 示意图, 流程图, 架构图, 画个图解释, 把这段内容做成图, 配图, 技术图, 可视化一段内容.
In this repository30 other files it gives its agents.
- AGENTS.mdAGENTS.md
- notion-im-helper/CLAUDE.mdCLAUDE.md
- 3d-printing-cad-designSkill
- agent-skills-syncSkill
- captain-resourceSkill
- csdn-publish-mcpSkill
- deep-research-report-pipelineSkill
- doc-plain-zhSkill
- feishu-bot-configSkill
- first-shotSkill
- investment-analysisSkill
- layoutfixSkill
- md2docx-formalSkill
- md-to-70asunflower-latexSkill
- multi-publish-mcpSkill
- notebook-patchingSkill
- notion-im-helperSkill
- paper-figuresSkill
- paper-reviewSkill
- paper-writingSkill
- project-initSkill
- project-reviewSkill
- siyuan-backup-syncSkill
- siyuan-kernel-opsSkill
- siyuan-markdown-formula-formatterSkill
- snapSkill
- strict-docx-auditSkill
- trek-mcp-opsSkill
- weekreportSkill
- xhs-pipelineSkill
The file
Files
visual-diagrams:把内容做成图
先选路径
| 要什么 | 走哪条 | 关键特征 |
| 信息卡 / 长图海报(文章要点、对比、清单) | ① HTML + CSS | 排版驱动,深色杂志风,导出 PNG |
| 示意图 / 流程图 / 电路图(每个元素都要精确控制) | ② 手写 SVG | 几何驱动,可交互,插进思源文档 |
| 技术架构图 / 系统图(要现成风格库,不想从零画) | ③ fireworks | 调用外部库,12 种风格一键切换 |
判断口径:内容为主、版式为辅 → ①;结构为主、要精确到坐标 → ②;要好看且不想调细节 → ③。
路径 ①:信息卡与长图
用 HTML/CSS 排版,Playwright 截图成 PNG。适合"把一篇文章/一段话变成一张能发出去的图"。
python scripts/screenshot.py card.html -o card.png --viewport 900x1200 --scale 2- 设计规范(字体层级、深色配色变量、卡片框架、移动端适配、布局模式)见
references/infographic-design.md - 文件命名用 kebab-case,英文优先
- 先输出一句分析「信息密度 高/中/低 → 采用 X 栏布局」,再出 HTML
路径 ②:手写 SVG 示意图
自己画 SVG,Chrome 渲染成 PNG,上传思源 assets 后插图。
python scripts/build_mini.py <图名> # 生成骨架
scripts/shoot.bat <svg> <png> # 无头 Chrome 渲染
python scripts/upload_swap_asset.py <png> # 上传思源并替换引用- 完整五步流程(绘制 → 渲染 → 上传 → 插图 → 校验刷新)与 MOSFET 电路符号语法见
references/svg-pipeline.md - 画完先自检:文字不压线、箭头接到框、标签居中——这些交给
layoutfix技能断言,别靠截图看
路径 ③:技术架构图(fireworks)
调用 fireworks-tech-graph 技能(第三方,物理在仓、git 不跟踪——见根目录 .gitignore 的 skillsync 区块)生成架构图、数据流、时序图、C4、云部署等。
- 12 种风格清单、命令模板、样式选择见
references/fireworks-styles.md - 质量契约与风格细则见
references/composition-quality-contract.md、references/style-1-flat-icon.md——风格规则真相源在../fireworks-tech-graph/references/, 本目录内同名文件是同步快照,改样式先改那边再同步过来 - 导出 PNG 用
scripts/png_export.js
通用约定
- 命名:kebab-case,英文优先,同一批图前缀一致(
fig-、card-、arch-) - 导出:PNG 至少 2 倍图;要印刷或放大时给 SVG/PDF 而非位图
- 校验:矢量图交付前跑
layoutfix check --strict,查字压图、超画布、箭头悬空 - 入库思源:图片走 assets 上传再引用,不要用外链
与 paper-figures 的边界
本技能做的是通用图解(信息卡、流程示意、架构图)。
论文里的数据图、多子图拼版、Graphical Abstract、学术海报归 paper-figures 技能,
那边有期刊版面预设和色盲质检,别在这里重复造。
自检
出图后按序确认:
- 文字有没有压到线或框上(
layoutfix check) - 导出尺寸够不够(PNG ≥ 2x,印刷要矢量)
- 文件名与目录符合命名约定
- 插进文档后引用路径对不对
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

