[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7ADAZ1WFR2865S104SADZMP":3,"skill-same:01M7ADAZ1WFR2865S104SADZMP":46},{"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":29,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":31,"install":32,"resources":45},"01M7ADAZ1WFR2865S104SADZMP","ai3d-gen",null,"Use when the user asks for generative 3D assets (text-to-3D, image-to-3D, mesh synthesis) via self-hosted open models. Routes to extensions\u002Fcomfyui-operator\u002Fwrapper.py against a local ComfyUI server, with direct-model fallbacks (TRELLIS, TripoSG, InstantMesh, Hunyuan3D). All free\u002Fopen weights, no paid SaaS.","Leostruka\u002Fdevin-bundle","Leostruka","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F81370685?v=4","01M7AD8MAK63VDNAWRDRRRW04N","leandro ostruka","01M7AD8MBS65YDS1KMMV7YPM07",0,"MIT","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Fskills\u002Fai3d-gen\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: ai3d-gen\ndescription: Use when the user asks for generative 3D assets (text-to-3D, image-to-3D, mesh synthesis) via self-hosted open models. Routes to extensions\u002Fcomfyui-operator\u002Fwrapper.py against a local ComfyUI server, with direct-model fallbacks (TRELLIS, TripoSG, InstantMesh, Hunyuan3D). All free\u002Fopen weights, no paid SaaS.\n---\n\n# AI 3D Generation (self-hosted)\n\nGenerate 3D assets locally via open-weight models. Two paths:\n\n1. **ComfyUI-3D-Pack** (primary): visual workflow server.\n   Wrapper: `extensions\u002Fcomfyui-operator\u002Fwrapper.py` (stdlib HTTP, no deps).\n2. **Direct model CLIs** (fallback \u002F batch): run the model repos' own\n   Python entrypoints, no ComfyUI needed.\n\nOutput of both paths: meshes (GLB\u002FOBJ\u002FPLY) that feed `operate-blender`\nfor cleanup, retopo, texturing, render, or `mesh-utils` for inspection.\n\n## Prerequisites\n\n- ComfyUI installed locally (user step): `git clone\n  https:\u002F\u002Fgithub.com\u002Fcomfyanonymous\u002FComfyUI && pip install -r\n  requirements.txt`, then `python main.py --port 8188`.\n- `ComfyUI-3D-Pack` custom node pack for full 3D graph coverage:\n  `git clone https:\u002F\u002Fgithub.com\u002FMrForExample\u002FComfyUI-3D-Pack\n  ComfyUI\u002Fcustom_nodes\u002FComfyUI-3D-Pack && pip install -r\n  ComfyUI\u002Fcustom_nodes\u002FComfyUI-3D-Pack\u002Frequirements.txt`\n- **3D-Pack requires NVIDIA GPU + CUDA toolkit (nvcc).** It eagerly\n  imports `nvdiffrast`, `spconv-cu126`, `nerfacc`, `slangtorch`; no CPU\n  wheels exist. On CPU-only machines the whole pack fails import\n  (ComfyUI core still runs). Use the direct-model fallback path there.\n- GPU strongly recommended for generation speed regardless (models are\n  diffusion transformers; CPU inference works but is very slow).\n\n## Wrapper commands\n\n```bash\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py status          # health + VRAM\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py submit wf.json  # queue job\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py history \u003Cpid>   # poll status + outputs\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py download \u003Cpid> --out dir\u002F\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py upload img.png  # input asset\n```\n\n`submit` requires **API-format JSON** (node-id -> {class_type, inputs}),\nthe shape produced by ComfyUI's \"Save (API Format)\". Build it\nprogrammatically or export from the UI once.\n\n`history` returns `files`: flattened output descriptors (filename,\nsubfolder, type) regardless of which node emitted them. `download`\nfetches each via `\u002Fview`.\n\n## Model choices (open weights, license-verified in research)\n\n| Model | Input | Quality | License | Where |\n|---|---|---|---|---|\n| TRELLIS | text\u002Fimg | high, GLB+texture | MIT | microsoft\u002FTRELLIS |\n| TripoSG | img | high mesh quality | MIT | VAST-AI\u002FTripoSG |\n| InstantMesh | img | fast | Apache-2.0 | TencentARC\u002FInstantMesh |\n| Hunyuan3D-2 | text\u002Fimg | high + PBR | Tencent permissive | Tencent\u002FHunyuan3D-2 |\n| Stable Fast 3D | img | fastest UV+mat | Stability Community | stabilityai\u002Fstable-fast-3d |\n\nPrefer TRELLIS\u002FTripoSG for permissive licensing; Hunyuan3D-2 when PBR\ntexture quality matters; Stable Fast 3D for speed.\n\n## Intake preflight (photo-sourced pipelines)\n\nBefore spending a reconstruction\u002Fsplat run on user photos:\n\n1. **Provenance**: source images should verify as real captures — run\n   `extensions\u002Fimage-forensics\u002Fverdict.py` per image when a camera\n   reference exists; synthetic inputs break PRNU assumptions and usually\n   produce bad splats.\n2. **Photometric QC**: run `extensions\u002Fppisp-prep\u002Fphoto_qc.py` on the\n   set; `correct`\u002F`recapture` verdicts route through the PPISP\n   correction pass or a re-shoot before generation.\n\n## Recipes\n\n**image -> textured GLB (ComfyUI path)**\n1. `wrapper.py upload photo.png`\n2. Build\u002Fsubmit a workflow JSON wiring `LoadImage` -> model node ->\n   `SaveMesh`\u002F`Preview3D` (node names per ComfyUI-3D-Pack version; query\n   `GET \u002Fobject_info` for exact class_type names on the live server).\n3. `wrapper.py submit wf.json`, poll `history`, `download` outputs.\n4. `python extensions\u002Fmesh-utils\u002Fmeshops.py info out.glb` to verify.\n5. Import into Blender via `operate-blender` for retopo\u002Fmaterial work.\n\n**text -> 3D without ComfyUI (direct)**\n- Clone the model repo, `pip install -r requirements.txt`, run its\n  inference script. E.g. TRELLIS example script produces GLB directly.\n  Then `mesh-utils`\u002F`operate-blender` downstream.\n\n## Boundaries\n\n- Never send user assets to hosted APIs (Meshy, Tripo cloud, etc.) unless\n  explicitly asked. The point of this stack is self-hosted.\n- ComfyUI not running -> wrapper exits 1 with a start hint. Don't spawn\n  the server silently; tell the user.\n- Check VRAM in `status` before queueing large models (>=8GB typical).\n\n## Model notes\n\n- ccore\u002FQwen3.5-4B-3D-Coder (ollama tag materion\u002Fqwen3.5-3d-coder:4b,\n  ~4B params, ~2.7GB Q4_K_M) generates interactive three.js scenes\n  in-browser from a prompt; a lightweight alternative to the ComfyUI\n  pipeline for quick 3D prototypes.\n","# AI 3D Generation (self-hosted)\n\nGenerate 3D assets locally via open-weight models. Two paths:\n\n1. **ComfyUI-3D-Pack** (primary): visual workflow server.\n   Wrapper: `extensions\u002Fcomfyui-operator\u002Fwrapper.py` (stdlib HTTP, no deps).\n2. **Direct model CLIs** (fallback \u002F batch): run the model repos' own\n   Python entrypoints, no ComfyUI needed.\n\nOutput of both paths: meshes (GLB\u002FOBJ\u002FPLY) that feed `operate-blender`\nfor cleanup, retopo, texturing, render, or `mesh-utils` for inspection.\n\n## Prerequisites\n\n- ComfyUI installed locally (user step): `git clone\n  https:\u002F\u002Fgithub.com\u002Fcomfyanonymous\u002FComfyUI && pip install -r\n  requirements.txt`, then `python main.py --port 8188`.\n- `ComfyUI-3D-Pack` custom node pack for full 3D graph coverage:\n  `git clone https:\u002F\u002Fgithub.com\u002FMrForExample\u002FComfyUI-3D-Pack\n  ComfyUI\u002Fcustom_nodes\u002FComfyUI-3D-Pack && pip install -r\n  ComfyUI\u002Fcustom_nodes\u002FComfyUI-3D-Pack\u002Frequirements.txt`\n- **3D-Pack requires NVIDIA GPU + CUDA toolkit (nvcc).** It eagerly\n  imports `nvdiffrast`, `spconv-cu126`, `nerfacc`, `slangtorch`; no CPU\n  wheels exist. On CPU-only machines the whole pack fails import\n  (ComfyUI core still runs). Use the direct-model fallback path there.\n- GPU strongly recommended for generation speed regardless (models are\n  diffusion transformers; CPU inference works but is very slow).\n\n## Wrapper commands\n\n```bash\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py status          # health + VRAM\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py submit wf.json  # queue job\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py history \u003Cpid>   # poll status + outputs\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py download \u003Cpid> --out dir\u002F\npython extensions\u002Fcomfyui-operator\u002Fwrapper.py upload img.png  # input asset\n```\n\n`submit` requires **API-format JSON** (node-id -> {class_type, inputs}),\nthe shape produced by ComfyUI's \"Save (API Format)\". Build it\nprogrammatically or export from the UI once.\n\n`history` returns `files`: flattened output descriptors (filename,\nsubfolder, type) regardless of which node emitted them. `download`\nfetches each via `\u002Fview`.\n\n## Model choices (open weights, license-verified in research)\n\n| Model | Input | Quality | License | Where |\n|---|---|---|---|---|\n| TRELLIS | text\u002Fimg | high, GLB+texture | MIT | microsoft\u002FTRELLIS |\n| TripoSG | img | high mesh quality | MIT | VAST-AI\u002FTripoSG |\n| InstantMesh | img | fast | Apache-2.0 | TencentARC\u002FInstantMesh |\n| Hunyuan3D-2 | text\u002Fimg | high + PBR | Tencent permissive | Tencent\u002FHunyuan3D-2 |\n| Stable Fast 3D | img | fastest UV+mat | Stability Community | stabilityai\u002Fstable-fast-3d |\n\nPrefer TRELLIS\u002FTripoSG for permissive licensing; Hunyuan3D-2 when PBR\ntexture quality matters; Stable Fast 3D for speed.\n\n## Intake preflight (photo-sourced pipelines)\n\nBefore spending a reconstruction\u002Fsplat run on user photos:\n\n1. **Provenance**: source images should verify as real captures — run\n   `extensions\u002Fimage-forensics\u002Fverdict.py` per image when a camera\n   reference exists; synthetic inputs break PRNU assumptions and usually\n   produce bad splats.\n2. **Photometric QC**: run `extensions\u002Fppisp-prep\u002Fphoto_qc.py` on the\n   set; `correct`\u002F`recapture` verdicts route through the PPISP\n   correction pass or a re-shoot before generation.\n\n## Recipes\n\n**image -> textured GLB (ComfyUI path)**\n1. `wrapper.py upload photo.png`\n2. Build\u002Fsubmit a workflow JSON wiring `LoadImage` -> model node ->\n   `SaveMesh`\u002F`Preview3D` (node names per ComfyUI-3D-Pack version; query\n   `GET \u002Fobject_info` for exact class_type names on the live server).\n3. `wrapper.py submit wf.json`, poll `history`, `download` outputs.\n4. `python extensions\u002Fmesh-utils\u002Fmeshops.py info out.glb` to verify.\n5. Import into Blender via `operate-blender` for retopo\u002Fmaterial work.\n\n**text -> 3D without ComfyUI (direct)**\n- Clone the model repo, `pip install -r requirements.txt`, run its\n  inference script. E.g. TRELLIS example script produces GLB directly.\n  Then `mesh-utils`\u002F`operate-blender` downstream.\n\n## Boundaries\n\n- Never send user assets to hosted APIs (Meshy, Tripo cloud, etc.) unless\n  explicitly asked. The point of this stack is self-hosted.\n- ComfyUI not running -> wrapper exits 1 with a start hint. Don't spawn\n  the server silently; tell the user.\n- Check VRAM in `status` before queueing large models (>=8GB typical).\n\n## Model notes\n\n- ccore\u002FQwen3.5-4B-3D-Coder (ollama tag materion\u002Fqwen3.5-3d-coder:4b,\n  ~4B params, ~2.7GB Q4_K_M) generates interactive three.js scenes\n  in-browser from a prompt; a lightweight alternative to the ComfyUI\n  pipeline for quick 3D prototypes.\n",{"name":5,"description":7},"main","02b95ddb62959a00b3c31442c2046aca0d19563c","skills\u002Fai3d-gen",1,"https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle","https:\u002F\u002Fleostruka.github.io\u002Fdevin-bundle\u002F","https:\u002F\u002Fgithub.com\u002FLeostruka","2026-11-15 05:19:58",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add Leostruka\u002Fdevin-bundle --skill ai3d-gen",{"tool":38,"command":39},"bunx","bunx skills add Leostruka\u002Fdevin-bundle --skill ai3d-gen",{"tool":41,"command":42},"pnpm","pnpm dlx skills add Leostruka\u002Fdevin-bundle --skill ai3d-gen","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Fskills\u002Fai3d-gen\u002FSKILL.md",true,[],[47,54,61,68,75,82],{"skillNo":48,"name":49,"nameZh":6,"description":50,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M7E0K3MG2XC4FSRV0MK6SFR1","montador-de-ideias","Use when assembling, expanding, or elevating a raw project idea or initial prompt into an ultra-detailed, first-principles dossier and compiled master prompt with journalistic OSINT market research, Tech Lead and Staff Architect justifications, atomic and modular hardening\u002Foptimization, and high-impact contact surfaces.","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Fskills\u002Fmontador-de-ideias\u002FSKILL.md",[],[],{"skillNo":55,"name":56,"nameZh":6,"description":57,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":60,"publishedAt":19},"01M7E0J7JJWZ8QMK01RZ3B3C9Z","git-github-adaptive-engineering","Use when any task touches git history, GitHub issues\u002FPRs\u002Freleases\u002FCI, repo governance, or branch and merge decisions. Runs the Fase Zero situational-awareness battery, picks one of five delivery flows via the adaptive decision engine, and enforces commit\u002FPR information-density standards. Orchestrates the gh, git-workflows, github-actions, pr, and debug\u002Ffinish skills.","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Fskills\u002Fgit-github-adaptive-engineering\u002FSKILL.md",[],[],{"skillNo":62,"name":63,"nameZh":6,"description":64,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M7E0GJFRPFEVQQMXEM58EWDM","adaptive-agenda-calendar-routine-engine","Use when planning or executing work across hours, days, weeks, or sprints. Turns the calendar into a strict time-blocked agenda (HH:MM-HH:MM blocks carrying a routine profile, expected inputs, and a handoff output artifact), runs the 5-stage cognitive flow (Radar and Hunt -> Triage plus Steve Jobs conception -> workbench manifest -> >=32-cycle loop -> Close plus Open-Minded Scrum), and unfolds macro calendar milestones into micro agenda blocks through extensions\u002Fgcal\u002Fwrapper.py.","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Fskills\u002Fadaptive-agenda-calendar-routine-engine\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":71,"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},"01M7AE2K9FTY0BVS48N0A9JFKF","demo","Use when demonstrating a skill in a fixture project.","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fdevin-config\u002Fproject\u002F.devin\u002Fskills\u002Fdemo\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},"01M7AE2G3DP38PTV3B1N7MW5TS","duplicate-skill","Use when testing duplicate skill name detection.","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fdevin-config\u002Fproject\u002F.devin\u002Fskills\u002Fbeta\u002FSKILL.md",[],[],{"skillNo":83,"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":84,"categories":85,"keywords":86,"publishedAt":19},"01M7AE2DBR8ZHT6T875MJ392GK","https:\u002F\u002Fgithub.com\u002FLeostruka\u002Fdevin-bundle\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fdevin-config\u002Fproject\u002F.devin\u002Fskills\u002Falpha\u002FSKILL.md",[],[]]