[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7F8ZJCY8F4P35VXE4XBFTGZ":3,"skill-same:01M7F8ZJCY8F4P35VXE4XBFTGZ":64},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M7F8ZJCY8F4P35VXE4XBFTGZ","xhs-product-image-design-report",null,"Use when the user provides one or more Xiaohongshu goods-detail links, product screenshots, or competitor product image folders and asks to collect public product fields\u002Fimages, compare main images\u002Fdetail images, explain why competitors designed them that way, and output a Chinese HTML product image design decision report for Xiaohongshu goods pages.","lza6\u002FArmor-piercing-system-prompts","lza6","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F128880206?v=4","01M7EB3K7WPX5T28EJ2NVHE6EG","Chinese-tingfeng","01M7EB3K97E1KQ16QWKJMFNGX8",11,"MIT","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: xhs-product-image-design-report\ndescription: Use when the user provides one or more Xiaohongshu goods-detail links, product screenshots, or competitor product image folders and asks to collect public product fields\u002Fimages, compare main images\u002Fdetail images, explain why competitors designed them that way, and output a Chinese HTML product image design decision report for Xiaohongshu goods pages.\n---\n\n# 小红书商品图设计决策报告\n\n## Goal\n\nTurn multiple Xiaohongshu product links or supplied product image folders into a design decision report answering:\n\n- 竞品主图\u002F详情图分别在解决什么购买问题。\n- 对方为什么这样设计。\n- 我的商品主图、副图、详情页长图应该怎么拍、怎么排、哪些表达需要证据。\n\nDefault output is a standalone Chinese HTML report matching the confirmed v10 report structure: local images, product links, comparison tables, one-product-per-row evidence wall, and a simulated vertical mobile detail-page layout.\n\n## Inputs\n\nAccept only Xiaohongshu product-analysis sources. Do not use this skill for Taobao, Tmall, JD, Douyin, Pinduoduo, Amazon, Shopify, or other platforms.\n\n- Xiaohongshu `https:\u002F\u002Fwww.xiaohongshu.com\u002Fgoods-detail\u002F...` URLs, preferably 3-10 links.\n- Exported Xiaohongshu product screenshots\u002Fimages for each product if live collection is blocked.\n- The user's own product资料: product name, category, target customer, price band, SKU, material, packaging, shipping\u002Fafter-sales rules.\n- The user's evidence: real product photos, supplier specs, test reports, material certificates, packaging photos, customer review excerpts, sales\u002Fclick\u002Fconversion data.\n\nIf own-product资料 or evidence is missing, continue with competitor analysis and mark assumptions\u002F证据缺口 clearly.\n\nWhen asking the user to provide screenshots manually, request these images for every Xiaohongshu product:\n\n1. 商品首屏截图: include product main image, title, price, sold count if visible, and shop\u002Fproduct header if visible.\n2. 商品图集截图 or saved carousel images: include all main images\u002Fsub-images in the product carousel, ideally 8-10 images per product.\n3. 详情页长图\u002F详情内容截图: scroll down and capture the product detail image area in order; use multiple screenshots if one long screenshot is not possible.\n4. SKU\u002Fspec screenshot: capture color, size, material, package, quantity, single\u002Fpair options, or variant selector if visible.\n5. 店铺\u002F口碑\u002F售后 screenshot: capture shop rating, shipping, return\u002Fafter-sales, packaging, delivery promises, or service badges if visible.\n\nTell the user that screenshots should preserve the original order. Do not crop away price, link\u002Ftitle context, or image captions that explain material\u002Fspecs.\n\n## Collection Capability And Dependencies\n\nCollection uses browser automation, not a Xiaohongshu official API.\n\nPreferred capability:\n\n- Playwright in a real browser session, through the local `playwright` skill\u002FCLI or a Node.js Playwright script.\n- Use existing logged-in browser state when public access is incomplete. If login\u002Fcaptcha\u002Fanti-bot blocks access, ask the user to open the links in their browser or provide screenshots\u002Fimages.\n\nRequired local dependencies for scripted collection:\n\n- Node.js and npm\u002Fnpx.\n- `playwright` package or the bundled Playwright CLI skill.\n- Network access to `xiaohongshu.com`.\n- A browser profile\u002Fsession that can view the goods pages when pages are not publicly accessible.\n\nRun the script with a Node.js environment that can resolve Playwright, for example:\n\n```bash\nnode scripts\u002Fcollect-xhs-goods.js --urls-file 输入资料\u002F商品链接.txt --out-dir 输出结果\n```\n\nWhen the host provides a bundled Node\u002FPlaywright runtime instead of a normal `node` command, use that host's documented runtime path. Do not copy a machine-specific absolute path from another user's environment.\n\nData that can usually be collected from visible public pages:\n\n- URL, page title, visible text.\n- Product title, shop name, price, sold count, rating\u002F口碑, SKU\u002Fspec text when visible.\n- `\u003Cimg>` image URLs, rendered image dimensions, alt text.\n- CSS background image URLs.\n- Full-page screenshots for evidence and fallback extraction.\n\nDo not claim access to private platform data, comments, backend metrics, order data, or hidden fields unless the user provides them.\n\n## Workflow\n\n### 1. Create Work Folder\n\nCreate a timestamped folder with this structure:\n\n```text\n小红书商品分析-{YYYYMMDD-HHMMSS}\u002F\n├── 输入资料\u002F\n│   ├── 商品链接.txt\n│   ├── 我的产品资料.md\n│   ├── 我的真实证据\u002F\n│   └── 竞品图片\u002F\n└── 输出结果\u002F\n    ├── 公开字段抽取.json\n    ├── 图片采集记录.md\n    ├── 竞品逐图拆解表.csv\n    ├── 小红书商品主图详情图设计决策分析报告.html\n    └── 截图验证\u002F\n```\n\n### 2. Collect Public Fields And Images\n\nUse `scripts\u002Fcollect-xhs-goods.js` as the starting point when collecting from live URLs. Provide `--urls-file` and `--out-dir` per task.\n\nFor each product:\n\n1. Open the URL with a mobile-like viewport first, because Xiaohongshu product pages are mobile-oriented.\n2. Wait for content and lazy images to load.\n3. Save full-page screenshot.\n4. Extract visible text, image URLs, background images, and rendered dimensions.\n5. Download or otherwise save usable product images into `输入资料\u002F竞品图片\u002F商品N_短标题\u002F`.\n6. Record failures and blocked fields in `图片采集记录.md`.\n\nIf live extraction is blocked, switch to user-provided screenshots\u002Fimages and explicitly label the source as \"用户提供图片\u002F截图\".\n\n### 3. Normalize Product Evidence\n\nCreate one normalized record per product:\n\n- 商品序号\n- 商品链接\n- 商品名称\n- 价格\u002F到手价\u002F划线价\n- 已售\u002F销量口径\n- 店铺名\u002F口碑\u002F粉丝数 if visible\n- SKU\u002Fspec\u002Fmaterial\u002Fpackaging if visible\n- Main image path\n- Detail image paths in display order\n- Notes about missing or uncertain fields\n\nKeep raw fields separate from inferred conclusions.\n\n### 4. Analyze Image Logic\n\nAnalyze each product across these dimensions:\n\n- 主图任务: click, style, first impression, product clarity.\n- 副图\u002F详情图任务: wearing effect, SKU choice, size, material, structure, packaging, after-sales, FAQ.\n- Purchase doubts solved: 好不好看, 能不能戴, 怎么选, 值不值, 是否真实, 收到什么.\n- Design patterns: 氛围实拍, 上耳效果, SKU矩阵, 结构说明, 礼物包装, 真实交付.\n- Evidence boundary: what can be said only with proof, e.g. 纯银, F136, 不过敏, 爆款, 久戴不痛.\n\nThe core judgment should not be \"which image looks better\"; it should be \"which purchase problem this image is solving\".\n\n### 5. Produce Design Recommendation\n\nOutput an actionable plan for the user's own images:\n\n- 主图 1+5 plan: 图序, 页面任务, 核心文案方向, 画面怎么设计, 需要素材\u002F证据, 设计判断.\n- 详情页 10 屏 order: first screen, style confirmation, doubts, structure, size, material, SKU, real delivery, after-sales, FAQ.\n- Execution checklist: 必须拍, 必须确认, 不要写.\n- Evidence gaps: what the user must provide before strong claims can be used.\n\n### 6. Build HTML Report\n\nUse a practical Chinese report UI:\n\n- Top summary with key conclusion metrics.\n- 商品级横向对比.\n- 主图\u002F详情图证据墙: follow the confirmed v10 layout, one product per row, product link visible, larger images than the earlier three-column wall.\n- 报告维度核对.\n- 对方为什么这样设计.\n- 竞品打法地图.\n- 我的商品图应该怎么设计.\n- 详情页模拟布局: follow the confirmed v10 behavior, phone frame with vertical scroll only, because users browse goods detail pages top-to-bottom; external cards should be a tidy two-row grid on desktop.\n- 执行清单.\n\nFor layout details, read `references\u002Freport-structure.md` when implementing the HTML.\n\n### 7. Verify\n\nAlways run lightweight checks before reporting completion:\n\n- Confirm the final HTML file exists.\n- Confirm every product link remains present in the report.\n- Confirm local image paths referenced by the report exist.\n- Confirm required sections are present: 商品对比, 图片证据墙, 设计原因, 设计方案, 详情页模拟布局, 执行清单.\n- Confirm the report states missing\u002Funcertain fields instead of treating inferred data as fact.\n\nRun Playwright\u002Fbrowser screenshot verification only when visual correctness matters, such as:\n\n- The user asks for a polished HTML report, UI redesign, or responsive layout check.\n- The report layout was newly created or substantially changed.\n- The detail-page phone mockup or image wall layout was modified.\n- The final artifact will be shared externally and needs visual QA.\n\nWhen running browser verification, start a local static server from the analysis folder, open the HTML with Playwright, capture desktop\u002Fmobile screenshots, and check image loading, product links, horizontal overflow, and vertical phone-frame scrolling. Console errors are acceptable only for unrelated favicon 404; mention this explicitly.\n\n## Output Naming\n\nUse a clear final filename:\n\n```text\n输出结果\u002F小红书商品主图详情图设计决策分析报告.html\n```\n\nIf iterating UI versions, suffix with `_v2`, `_v3`, etc., and tell the user which version is recommended.\n","# 小红书商品图设计决策报告\n\n## Goal\n\nTurn multiple Xiaohongshu product links or supplied product image folders into a design decision report answering:\n\n- 竞品主图\u002F详情图分别在解决什么购买问题。\n- 对方为什么这样设计。\n- 我的商品主图、副图、详情页长图应该怎么拍、怎么排、哪些表达需要证据。\n\nDefault output is a standalone Chinese HTML report matching the confirmed v10 report structure: local images, product links, comparison tables, one-product-per-row evidence wall, and a simulated vertical mobile detail-page layout.\n\n## Inputs\n\nAccept only Xiaohongshu product-analysis sources. Do not use this skill for Taobao, Tmall, JD, Douyin, Pinduoduo, Amazon, Shopify, or other platforms.\n\n- Xiaohongshu `https:\u002F\u002Fwww.xiaohongshu.com\u002Fgoods-detail\u002F...` URLs, preferably 3-10 links.\n- Exported Xiaohongshu product screenshots\u002Fimages for each product if live collection is blocked.\n- The user's own product资料: product name, category, target customer, price band, SKU, material, packaging, shipping\u002Fafter-sales rules.\n- The user's evidence: real product photos, supplier specs, test reports, material certificates, packaging photos, customer review excerpts, sales\u002Fclick\u002Fconversion data.\n\nIf own-product资料 or evidence is missing, continue with competitor analysis and mark assumptions\u002F证据缺口 clearly.\n\nWhen asking the user to provide screenshots manually, request these images for every Xiaohongshu product:\n\n1. 商品首屏截图: include product main image, title, price, sold count if visible, and shop\u002Fproduct header if visible.\n2. 商品图集截图 or saved carousel images: include all main images\u002Fsub-images in the product carousel, ideally 8-10 images per product.\n3. 详情页长图\u002F详情内容截图: scroll down and capture the product detail image area in order; use multiple screenshots if one long screenshot is not possible.\n4. SKU\u002Fspec screenshot: capture color, size, material, package, quantity, single\u002Fpair options, or variant selector if visible.\n5. 店铺\u002F口碑\u002F售后 screenshot: capture shop rating, shipping, return\u002Fafter-sales, packaging, delivery promises, or service badges if visible.\n\nTell the user that screenshots should preserve the original order. Do not crop away price, link\u002Ftitle context, or image captions that explain material\u002Fspecs.\n\n## Collection Capability And Dependencies\n\nCollection uses browser automation, not a Xiaohongshu official API.\n\nPreferred capability:\n\n- Playwright in a real browser session, through the local `playwright` skill\u002FCLI or a Node.js Playwright script.\n- Use existing logged-in browser state when public access is incomplete. If login\u002Fcaptcha\u002Fanti-bot blocks access, ask the user to open the links in their browser or provide screenshots\u002Fimages.\n\nRequired local dependencies for scripted collection:\n\n- Node.js and npm\u002Fnpx.\n- `playwright` package or the bundled Playwright CLI skill.\n- Network access to `xiaohongshu.com`.\n- A browser profile\u002Fsession that can view the goods pages when pages are not publicly accessible.\n\nRun the script with a Node.js environment that can resolve Playwright, for example:\n\n```bash\nnode scripts\u002Fcollect-xhs-goods.js --urls-file 输入资料\u002F商品链接.txt --out-dir 输出结果\n```\n\nWhen the host provides a bundled Node\u002FPlaywright runtime instead of a normal `node` command, use that host's documented runtime path. Do not copy a machine-specific absolute path from another user's environment.\n\nData that can usually be collected from visible public pages:\n\n- URL, page title, visible text.\n- Product title, shop name, price, sold count, rating\u002F口碑, SKU\u002Fspec text when visible.\n- `\u003Cimg>` image URLs, rendered image dimensions, alt text.\n- CSS background image URLs.\n- Full-page screenshots for evidence and fallback extraction.\n\nDo not claim access to private platform data, comments, backend metrics, order data, or hidden fields unless the user provides them.\n\n## Workflow\n\n### 1. Create Work Folder\n\nCreate a timestamped folder with this structure:\n\n```text\n小红书商品分析-{YYYYMMDD-HHMMSS}\u002F\n├── 输入资料\u002F\n│   ├── 商品链接.txt\n│   ├── 我的产品资料.md\n│   ├── 我的真实证据\u002F\n│   └── 竞品图片\u002F\n└── 输出结果\u002F\n    ├── 公开字段抽取.json\n    ├── 图片采集记录.md\n    ├── 竞品逐图拆解表.csv\n    ├── 小红书商品主图详情图设计决策分析报告.html\n    └── 截图验证\u002F\n```\n\n### 2. Collect Public Fields And Images\n\nUse `scripts\u002Fcollect-xhs-goods.js` as the starting point when collecting from live URLs. Provide `--urls-file` and `--out-dir` per task.\n\nFor each product:\n\n1. Open the URL with a mobile-like viewport first, because Xiaohongshu product pages are mobile-oriented.\n2. Wait for content and lazy images to load.\n3. Save full-page screenshot.\n4. Extract visible text, image URLs, background images, and rendered dimensions.\n5. Download or otherwise save usable product images into `输入资料\u002F竞品图片\u002F商品N_短标题\u002F`.\n6. Record failures and blocked fields in `图片采集记录.md`.\n\nIf live extraction is blocked, switch to user-provided screenshots\u002Fimages and explicitly label the source as \"用户提供图片\u002F截图\".\n\n### 3. Normalize Product Evidence\n\nCreate one normalized record per product:\n\n- 商品序号\n- 商品链接\n- 商品名称\n- 价格\u002F到手价\u002F划线价\n- 已售\u002F销量口径\n- 店铺名\u002F口碑\u002F粉丝数 if visible\n- SKU\u002Fspec\u002Fmaterial\u002Fpackaging if visible\n- Main image path\n- Detail image paths in display order\n- Notes about missing or uncertain fields\n\nKeep raw fields separate from inferred conclusions.\n\n### 4. Analyze Image Logic\n\nAnalyze each product across these dimensions:\n\n- 主图任务: click, style, first impression, product clarity.\n- 副图\u002F详情图任务: wearing effect, SKU choice, size, material, structure, packaging, after-sales, FAQ.\n- Purchase doubts solved: 好不好看, 能不能戴, 怎么选, 值不值, 是否真实, 收到什么.\n- Design patterns: 氛围实拍, 上耳效果, SKU矩阵, 结构说明, 礼物包装, 真实交付.\n- Evidence boundary: what can be said only with proof, e.g. 纯银, F136, 不过敏, 爆款, 久戴不痛.\n\nThe core judgment should not be \"which image looks better\"; it should be \"which purchase problem this image is solving\".\n\n### 5. Produce Design Recommendation\n\nOutput an actionable plan for the user's own images:\n\n- 主图 1+5 plan: 图序, 页面任务, 核心文案方向, 画面怎么设计, 需要素材\u002F证据, 设计判断.\n- 详情页 10 屏 order: first screen, style confirmation, doubts, structure, size, material, SKU, real delivery, after-sales, FAQ.\n- Execution checklist: 必须拍, 必须确认, 不要写.\n- Evidence gaps: what the user must provide before strong claims can be used.\n\n### 6. Build HTML Report\n\nUse a practical Chinese report UI:\n\n- Top summary with key conclusion metrics.\n- 商品级横向对比.\n- 主图\u002F详情图证据墙: follow the confirmed v10 layout, one product per row, product link visible, larger images than the earlier three-column wall.\n- 报告维度核对.\n- 对方为什么这样设计.\n- 竞品打法地图.\n- 我的商品图应该怎么设计.\n- 详情页模拟布局: follow the confirmed v10 behavior, phone frame with vertical scroll only, because users browse goods detail pages top-to-bottom; external cards should be a tidy two-row grid on desktop.\n- 执行清单.\n\nFor layout details, read `references\u002Freport-structure.md` when implementing the HTML.\n\n### 7. Verify\n\nAlways run lightweight checks before reporting completion:\n\n- Confirm the final HTML file exists.\n- Confirm every product link remains present in the report.\n- Confirm local image paths referenced by the report exist.\n- Confirm required sections are present: 商品对比, 图片证据墙, 设计原因, 设计方案, 详情页模拟布局, 执行清单.\n- Confirm the report states missing\u002Funcertain fields instead of treating inferred data as fact.\n\nRun Playwright\u002Fbrowser screenshot verification only when visual correctness matters, such as:\n\n- The user asks for a polished HTML report, UI redesign, or responsive layout check.\n- The report layout was newly created or substantially changed.\n- The detail-page phone mockup or image wall layout was modified.\n- The final artifact will be shared externally and needs visual QA.\n\nWhen running browser verification, start a local static server from the analysis folder, open the HTML with Playwright, capture desktop\u002Fmobile screenshots, and check image loading, product links, horizontal overflow, and vertical phone-frame scrolling. Console errors are acceptable only for unrelated favicon 404; mention this explicitly.\n\n## Output Naming\n\nUse a clear final filename:\n\n```text\n输出结果\u002F小红书商品主图详情图设计决策分析报告.html\n```\n\nIf iterating UI versions, suffix with `_v2`, `_v3`, etc., and tell the user which version is recommended.\n",{"name":5,"description":7},"main","d69096e239f2a89aabc2cd078c47664a9d3b1e11","06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report",3,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts","https:\u002F\u002Fgithub.com\u002Flza6","个人vx：Tf00798\r\n个人✈：@TFGY999","2026-10-08 14:29:22",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add lza6\u002FArmor-piercing-system-prompts --skill xhs-product-image-design-report",{"tool":38,"command":39},"bunx","bunx skills add lza6\u002FArmor-piercing-system-prompts --skill xhs-product-image-design-report",{"tool":41,"command":42},"pnpm","pnpm dlx skills add lza6\u002FArmor-piercing-system-prompts --skill xhs-product-image-design-report","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report\u002FSKILL.md",true,[46,52,56,60],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"AGENTS.md","REFERENCE","参考文档",1454,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report\u002FAGENTS.md",{"path":53,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":54,"sourceHtmlUrl":55},"agents\u002Fopenai.yaml",294,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report\u002Fagents\u002Fopenai.yaml",{"path":57,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":58,"sourceHtmlUrl":59},"README.md",8946,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report\u002FREADME.md",{"path":61,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":62,"sourceHtmlUrl":63},"references\u002Freport-structure.md",2381,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fduuhuang-ai__xhs-product-image-design-report\u002Freferences\u002Freport-structure.md",[65,72,77,84,91,98],{"skillNo":66,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7FFG485KY83TD6HSC4YPCMC","obsidian-clipper-template-creator","Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator\u002FSKILL.md",[],[],{"skillNo":73,"name":67,"nameZh":6,"description":68,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7FFG0T6AZ0STEJPSB3CX84K","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobsidian-clipper-template-creator-antigravity-awesome-skills-main\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7FFFY0QEM14SKWPDYGMXM2P","python-api","Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP\u002FmAP semantics, and source-style import diagnostics.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Fpython-api\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7FFFVZQZ2GJ0C0TGP7JHD20","file-evaluation","Folder-based PASCAL VOC AP\u002FmAP evaluation from ground-truth and detection text files.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002Fsub-skills\u002Ffile-evaluation\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M7FFFSFNYSE77SJ6PE1CK2DA","object-detection-metrics","Guides legacy Object-Detection-Metrics PASCAL VOC AP\u002FmAP evaluation from detection text folders and direct Python API classes.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fobject-detection-metrics\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7FFFP0GCK1RVXBCPQ8APBMT","nx-workspace-patterns","Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.","https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F07-后端与架构\u002Fnx-workspace-patterns\u002FSKILL.md",[],[]]