[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7FC7GDBQZTMXQ97HT6CQH1T":3,"skill-same:01M7FC7GDBQZTMXQ97HT6CQH1T":62},{"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":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":33,"updatedAt":34,"install":35,"resources":47},"01M7FC7GDBQZTMXQ97HT6CQH1T","visualization-and-serving",null,"Guides the historical SECOND KITTI web viewer, its Flask API, browser configuration, visualization payloads, and guarded checkpoint inference routes.","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\u002Fsecond-pytorch\u002Fsub-skills\u002Fvisualization-and-serving\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: visualization-and-serving\ndescription: \"Guides the historical SECOND KITTI web viewer, its Flask API,\n  browser configuration, visualization payloads, and guarded checkpoint\n  inference routes.\"\ndisable-model-invocation: true\nmetadata:\n  disco-role: operating\nlicense: MIT\n---\n\n# Visualization and serving\n\nUse this route for requests such as **start the KITTI viewer web backend**,\n**serve the browser UI**, **load point clouds or detections**, **configure\nFlask\u002FCORS**, **buildNet**, **run viewer inference**, or **use the PyQt viewer**.\nThe supported route here is the old repository's **web KITTI lidar viewer**;\nthe desktop Qt viewer is historical and deprecated.\n\n## Safety and support boundary\n\n- This is a source-distilled protocol, not a replacement server. A server is\n  started only when the user explicitly asks for it, with user-owned paths and\n  ports. No server, browser, GUI, dataset, or detector execution is part of\n  skill verification.\n- The checkout has no setup metadata. Do not assume `pip install -e .` works.\n  First run the safe dependency probe:\n\n  ```bash\n  python sub-skills\u002Fvisualization-and-serving\u002Fscripts\u002Fcheck_viewer_deps.py --json\n  ```\n\n  Missing components are reported without launching Flask. `--help` is also\n  safe. A passing import probe does **not** prove detector compatibility.\n- The historical model path uses legacy spconv\u002FNumba APIs. Current spconv 2.x\n  is not proven compatible: in the inspected environment the legacy\n  `spconv.utils.VoxelGeneratorV2` and `non_max_suppression` symbols are absent.\n  Do not claim that `buildNet` or inference works, even if CUDA and PyTorch\n  import successfully. Route compatibility and checkpoints to\n  [training-and-inference](..\u002Ftraining-and-inference\u002FSKILL.md).\n- The backend binds to `127.0.0.1` and enables permissive Flask-CORS. CORS does\n  not make a loopback server remotely reachable, and the API has no\n  authentication. Keep it local or use an explicitly user-controlled secure\n  tunnel; never expose arbitrary host paths or a public service by default.\n\n## Web workflow\n\n1. Prepare a local KITTI-style dataset and generated info pickle. Use\n   [data-preparation](..\u002Fdata-preparation\u002FSKILL.md) for layout and info\n   validation; the browser sends filesystem paths to the **backend host**, not\n   to the browser machine.\n2. In a caller-managed environment where the `second` package and web\n   dependencies are importable, start the backend only on request:\n\n   ```bash\n   python -m second.kittiviewer.backend.main main --port=\u003CBACKEND_PORT>\n   ```\n\n   The source default is `16666`; the command is a long-running foreground\n   service. Stop it with the terminal's normal interrupt. A safe import\u002Fhelp\n   diagnostic is:\n\n   ```bash\n   python -m second.kittiviewer.backend.main --help\n   ```\n\n   If import fails before help, stop and use\n   [troubleshooting](references\u002Ftroubleshooting.md); do not silently substitute\n   modern detector APIs.\n3. Serve the supplied web frontend assets separately, from the directory that\n   contains `index.html`, without exposing a broad parent directory:\n\n   ```bash\n   python -m http.server \u003CFRONTEND_PORT> --directory \u003Cfrontend-assets-directory>\n   ```\n\n   The README's example uses port `8000`. Open\n   `http:\u002F\u002F127.0.0.1:\u003CFRONTEND_PORT>\u002F` in a browser and set the UI's **backend**\n   field to `http:\u002F\u002F127.0.0.1:\u003CBACKEND_PORT>` (the UI adds `http:\u002F\u002F` when a\n   scheme is omitted). Verify browser developer tools show POSTs to the same\n   backend origin and no CORS error.\n4. Set **datasetClassName** to the registered class, normally\n   `KittiDataset`, and set `rootPath` and `infoPath` to paths readable by the\n   backend. Click **load**. Expect `status: normal` and an `image_indexes` list.\n   Enter an index shown by that list in the bottom control and press Enter;\n   previous\u002Fnext only works after a successful load.\n5. Optionally set **detPath** and click **loadDet**. The backend accepts either\n   a pickled detection result file or a directory in the historical KITTI label\n   format. Treat this as best-effort legacy state: the current endpoint stores\n   `dt_annos`, but the point-cloud response does not consistently merge those\n   annotations into its payload. Check the actual response before relying on\n   `drawDet`.\n6. For checkpoint inference, first complete **load**, then set\n   **checkpointPath** and **configPath**, click **buildNet**, and only then\n   click **inference**. This route is guarded and unverified for modern spconv;\n   use [training-and-inference](..\u002Ftraining-and-inference\u002FSKILL.md) for\n   checkpoint\u002Fconfig compatibility and [api-reference](references\u002Fapi-reference.md)\n   for exact request fields.\n\n## Routes and handoffs\n\n- Read [web-viewer.md](references\u002Fweb-viewer.md) for the browser\u002Fbackend state\n  sequence, payload encodings, and visualization behavior.\n- Read [api-reference.md](references\u002Fapi-reference.md) before constructing a\n  direct HTTP request or diagnosing a response shape.\n- Read [troubleshooting.md](references\u002Ftroubleshooting.md) for missing imports,\n  CORS\u002FURL mismatches, invalid info or detection inputs, API ordering errors,\n  and legacy checkpoint failures.\n- Route dataset directories, info generation, and dataset class registration\n  to [data-preparation](..\u002Fdata-preparation\u002FSKILL.md). Route box coordinate\n  conventions and evaluation-format conversion to\n  [geometry-and-evaluation](..\u002Fgeometry-and-evaluation\u002FSKILL.md).\n\n## Explicit non-goals\n\nDo not copy, wrap, or launch the deprecated `viewer.py`, `glwidget.py`, or\n`control_panel.py` desktop application. Those Qt\u002FOpenGL modules require a\nseparate GUI stack and are retained only as deprecation evidence. Prefer the\nweb lidar path, and recommend a maintained successor such as OpenPCDet or\nMMDetection3D for new work.\n","# Visualization and serving\n\nUse this route for requests such as **start the KITTI viewer web backend**,\n**serve the browser UI**, **load point clouds or detections**, **configure\nFlask\u002FCORS**, **buildNet**, **run viewer inference**, or **use the PyQt viewer**.\nThe supported route here is the old repository's **web KITTI lidar viewer**;\nthe desktop Qt viewer is historical and deprecated.\n\n## Safety and support boundary\n\n- This is a source-distilled protocol, not a replacement server. A server is\n  started only when the user explicitly asks for it, with user-owned paths and\n  ports. No server, browser, GUI, dataset, or detector execution is part of\n  skill verification.\n- The checkout has no setup metadata. Do not assume `pip install -e .` works.\n  First run the safe dependency probe:\n\n  ```bash\n  python sub-skills\u002Fvisualization-and-serving\u002Fscripts\u002Fcheck_viewer_deps.py --json\n  ```\n\n  Missing components are reported without launching Flask. `--help` is also\n  safe. A passing import probe does **not** prove detector compatibility.\n- The historical model path uses legacy spconv\u002FNumba APIs. Current spconv 2.x\n  is not proven compatible: in the inspected environment the legacy\n  `spconv.utils.VoxelGeneratorV2` and `non_max_suppression` symbols are absent.\n  Do not claim that `buildNet` or inference works, even if CUDA and PyTorch\n  import successfully. Route compatibility and checkpoints to\n  [training-and-inference](..\u002Ftraining-and-inference\u002FSKILL.md).\n- The backend binds to `127.0.0.1` and enables permissive Flask-CORS. CORS does\n  not make a loopback server remotely reachable, and the API has no\n  authentication. Keep it local or use an explicitly user-controlled secure\n  tunnel; never expose arbitrary host paths or a public service by default.\n\n## Web workflow\n\n1. Prepare a local KITTI-style dataset and generated info pickle. Use\n   [data-preparation](..\u002Fdata-preparation\u002FSKILL.md) for layout and info\n   validation; the browser sends filesystem paths to the **backend host**, not\n   to the browser machine.\n2. In a caller-managed environment where the `second` package and web\n   dependencies are importable, start the backend only on request:\n\n   ```bash\n   python -m second.kittiviewer.backend.main main --port=\u003CBACKEND_PORT>\n   ```\n\n   The source default is `16666`; the command is a long-running foreground\n   service. Stop it with the terminal's normal interrupt. A safe import\u002Fhelp\n   diagnostic is:\n\n   ```bash\n   python -m second.kittiviewer.backend.main --help\n   ```\n\n   If import fails before help, stop and use\n   [troubleshooting](references\u002Ftroubleshooting.md); do not silently substitute\n   modern detector APIs.\n3. Serve the supplied web frontend assets separately, from the directory that\n   contains `index.html`, without exposing a broad parent directory:\n\n   ```bash\n   python -m http.server \u003CFRONTEND_PORT> --directory \u003Cfrontend-assets-directory>\n   ```\n\n   The README's example uses port `8000`. Open\n   `http:\u002F\u002F127.0.0.1:\u003CFRONTEND_PORT>\u002F` in a browser and set the UI's **backend**\n   field to `http:\u002F\u002F127.0.0.1:\u003CBACKEND_PORT>` (the UI adds `http:\u002F\u002F` when a\n   scheme is omitted). Verify browser developer tools show POSTs to the same\n   backend origin and no CORS error.\n4. Set **datasetClassName** to the registered class, normally\n   `KittiDataset`, and set `rootPath` and `infoPath` to paths readable by the\n   backend. Click **load**. Expect `status: normal` and an `image_indexes` list.\n   Enter an index shown by that list in the bottom control and press Enter;\n   previous\u002Fnext only works after a successful load.\n5. Optionally set **detPath** and click **loadDet**. The backend accepts either\n   a pickled detection result file or a directory in the historical KITTI label\n   format. Treat this as best-effort legacy state: the current endpoint stores\n   `dt_annos`, but the point-cloud response does not consistently merge those\n   annotations into its payload. Check the actual response before relying on\n   `drawDet`.\n6. For checkpoint inference, first complete **load**, then set\n   **checkpointPath** and **configPath**, click **buildNet**, and only then\n   click **inference**. This route is guarded and unverified for modern spconv;\n   use [training-and-inference](..\u002Ftraining-and-inference\u002FSKILL.md) for\n   checkpoint\u002Fconfig compatibility and [api-reference](references\u002Fapi-reference.md)\n   for exact request fields.\n\n## Routes and handoffs\n\n- Read [web-viewer.md](references\u002Fweb-viewer.md) for the browser\u002Fbackend state\n  sequence, payload encodings, and visualization behavior.\n- Read [api-reference.md](references\u002Fapi-reference.md) before constructing a\n  direct HTTP request or diagnosing a response shape.\n- Read [troubleshooting.md](references\u002Ftroubleshooting.md) for missing imports,\n  CORS\u002FURL mismatches, invalid info or detection inputs, API ordering errors,\n  and legacy checkpoint failures.\n- Route dataset directories, info generation, and dataset class registration\n  to [data-preparation](..\u002Fdata-preparation\u002FSKILL.md). Route box coordinate\n  conventions and evaluation-format conversion to\n  [geometry-and-evaluation](..\u002Fgeometry-and-evaluation\u002FSKILL.md).\n\n## Explicit non-goals\n\nDo not copy, wrap, or launch the deprecated `viewer.py`, `glwidget.py`, or\n`control_panel.py` desktop application. Those Qt\u002FOpenGL modules require a\nseparate GUI stack and are retained only as deprecation evidence. Prefer the\nweb lidar path, and recommend a maintained successor such as OpenPCDet or\nMMDetection3D for new work.\n",{"name":5,"license":15,"metadata":24,"description":7,"disable-model-invocation":26},{"disco-role":25},"operating",true,"main","d69096e239f2a89aabc2cd078c47664a9d3b1e11","06-前端与设计\u002Fsecond-pytorch\u002Fsub-skills\u002Fvisualization-and-serving",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":36,"prompt":46,"zipAvailable":26},[37,40,43],{"tool":38,"command":39},"npx","npx skills add lza6\u002FArmor-piercing-system-prompts --skill visualization-and-serving",{"tool":41,"command":42},"bunx","bunx skills add lza6\u002FArmor-piercing-system-prompts --skill visualization-and-serving",{"tool":44,"command":45},"pnpm","pnpm dlx skills add lza6\u002FArmor-piercing-system-prompts --skill visualization-and-serving","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\u002Fsecond-pytorch\u002Fsub-skills\u002Fvisualization-and-serving\u002FSKILL.md",[48,54,58],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fapi-reference.md","REFERENCE","参考文档",8321,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fsecond-pytorch\u002Fsub-skills\u002Fvisualization-and-serving\u002Freferences\u002Fapi-reference.md",{"path":55,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Ftroubleshooting.md",10425,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fsecond-pytorch\u002Fsub-skills\u002Fvisualization-and-serving\u002Freferences\u002Ftroubleshooting.md",{"path":59,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Fweb-viewer.md",9238,"https:\u002F\u002Fgithub.com\u002Flza6\u002FArmor-piercing-system-prompts\u002Fblob\u002Fmain\u002F06-前端与设计\u002Fsecond-pytorch\u002Fsub-skills\u002Fvisualization-and-serving\u002Freferences\u002Fweb-viewer.md",[63,70,75,82,89,96],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"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":71,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":72,"categories":73,"keywords":74,"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":76,"name":77,"nameZh":6,"description":78,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"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":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"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":90,"name":91,"nameZh":6,"description":92,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":93,"categories":94,"keywords":95,"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":97,"name":98,"nameZh":6,"description":99,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"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",[],[]]