[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6GDSE23EVP8AMQ081P3DX6G":3,"skill-same:01M6GDSE23EVP8AMQ081P3DX6G":122},{"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":34,"publishedAt":55,"linkOnly":56,"markdownContent":57,"markdownBody":58,"frontMatter":59,"refName":100,"commitSha":101,"skillDirectory":102,"forksCount":103,"repositoryHtmlUrl":104,"repositoryHomepageUrl":6,"ownerHtmlUrl":105,"ownerBlogUrl":6,"ownerDescription":106,"updatedAt":107,"install":108,"resources":121},"01M6GDSE23EVP8AMQ081P3DX6G","kanban-video-orchestrator",null,"Plan, set up, and monitor a multi-agent video production pipeline backed by Hermes Kanban. Use when the user wants to make ANY video — narrative film, product\u002Fmarketing, music video, explainer, ASCII\u002Fterminal art, abstract\u002Fgenerative loop, comic, 3D, real-time\u002Finstallation — and the work warrants decomposition into specialized profiles (writer, designer, animator, renderer, voice, editor, etc.) coordinated through a kanban board. Performs adaptive discovery to scope the brief, designs an appropriate team for the requested style, generates the setup script that creates Hermes profiles + initial kanban task, then helps monitor execution and intervene when tasks stall or fail. Routes scenes to whichever Hermes rendering \u002F audio \u002F design skill fits each beat (`ascii-video`, `manim-video`, `p5js`, `comfyui`, `touchdesigner-mcp`, `blender-mcp`, `pixel-art`, `baoyu-comic`, `claude-design`, `excalidraw`, `songsee`, `heartmula`, …) plus external APIs for TTS, image-gen, and image-to-video as needed.","zapabob\u002FSkills","zapabob","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F130830376?v=4","01M6DBMYZAGRTFQDSQYW3W764P","峯岸 亮","01M6DBMZ3GKDSJ01NZ9XTZD7JD",2,"MIT","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002Fbe\u002Fbe54d5f681fff01e2ce3b4684d9644382343f8e12a212c6ca482b90a654c41f6\u002FSKILL.md",[18,23,26,29],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000015","视觉设计","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体",{"categoryNo":27,"categoryName":28,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":30,"categoryName":31,"taxonomyType":32,"taxonomyTypeDesc":33},"01JBSEEDCAT000000000000032","设计师","PROFESSION","适用职业",[35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54],"kanban video orchestrator","Video","kanban","orchestrator","specialized","pixel-art","SOUL.md","worker","project","time","writer","design","loop","team","Audio","3D","search","architecture","image","MCP","2026-11-10 10:12:30",false,"---\r\nname: kanban-video-orchestrator\r\ndescription: Plan, set up, and monitor a multi-agent video production pipeline backed by Hermes Kanban. Use when the user wants to make ANY video — narrative film, product\u002Fmarketing, music video, explainer, ASCII\u002Fterminal art, abstract\u002Fgenerative loop, comic, 3D, real-time\u002Finstallation — and the work warrants decomposition into specialized profiles (writer, designer, animator, renderer, voice, editor, etc.) coordinated through a kanban board. Performs adaptive discovery to scope the brief, designs an appropriate team for the requested style, generates the setup script that creates Hermes profiles + initial kanban task, then helps monitor execution and intervene when tasks stall or fail. Routes scenes to whichever Hermes rendering \u002F audio \u002F design skill fits each beat (`ascii-video`, `manim-video`, `p5js`, `comfyui`, `touchdesigner-mcp`, `blender-mcp`, `pixel-art`, `baoyu-comic`, `claude-design`, `excalidraw`, `songsee`, `heartmula`, …) plus external APIs for TTS, image-gen, and image-to-video as needed.\r\nversion: 1.0.0\r\nauthor: [SHL0MS, alt-glitch]\r\nlicense: MIT\r\nplatforms: [linux, macos, windows]\r\nmetadata:\r\n  hermes:\r\n    tags: [video, kanban, multi-agent, orchestration, production-pipeline]\r\n    related_skills: [kanban-orchestrator, kanban-worker, ascii-video, manim-video, p5js, comfyui, touchdesigner-mcp, blender-mcp, pixel-art, ascii-art, songwriting-and-ai-music, heartmula, songsee, spotify, youtube-content, claude-design, excalidraw, architecture-diagram, concept-diagrams, baoyu-comic, baoyu-infographic, humanizer, gif-search, meme-generation]\r\n    credits: |\r\n      The single-project workspace layout, profile-config patching pattern,\r\n      SOUL.md-per-profile model, TEAM.md task-graph convention, and\r\n      `--workspace dir:\u003Cpath>` discipline are adapted from alt-glitch's\r\n      original multi-agent video pipeline at\r\n      https:\u002F\u002Fgithub.com\u002FNousResearch\u002Fkanban-video-pipeline.\r\n---\r\n\r\n# Kanban Video Orchestrator\r\n\r\nWrap any video request — from a 15-second product teaser to a 5-minute narrative\r\nshort to a music video to an ASCII loop — in a Hermes Kanban pipeline that\r\ndecomposes the work to specialized agent profiles.\r\n\r\nThis skill does **not** render anything itself. It is a meta-pipeline that:\r\n\r\n1. **Scopes** the request through targeted discovery\r\n2. **Designs** an appropriate team (which roles, which tools per role) based on the style\r\n3. **Generates** a setup script that creates Hermes profiles, project workspace, and the initial kanban task\r\n4. **Hands off** to the director profile, which decomposes via the kanban\r\n5. **Monitors** execution, helps intervene when tasks stall or fail\r\n\r\nThe actual rendering happens inside the kanban once it's running, via whichever\r\nexisting skills + tools fit the scenes — `ascii-video`, `manim-video`, `p5js`,\r\n`comfyui`, `touchdesigner-mcp`, `blender-mcp`, `songwriting-and-ai-music`,\r\n`heartmula`, external APIs, or plain Python with PIL + ffmpeg.\r\n\r\n## When NOT to use this skill\r\n\r\n- The video is one continuous procedural project that needs no specialists. Just write the code directly.\r\n- The user wants a quick one-shot conversion (e.g. \"convert this mp4 to a GIF\") — use ffmpeg directly.\r\n- The output is a static image, GIF, or audio-only artifact — use the matching specific skill (`ascii-art`, `gifs`, `meme-generation`, `songwriting-and-ai-music`).\r\n- The work fits a single existing skill cleanly (e.g. a pure ASCII video — just use `ascii-video`).\r\n\r\n## Workflow\r\n\r\n```\r\nDISCOVER  →  BRIEF  →  TEAM DESIGN  →  SETUP  →  EXECUTE  →  MONITOR\r\n```\r\n\r\n### Step 1 — Discover (ask the right questions)\r\n\r\nThe discovery process is **adaptive**: ask only what is actually needed. Always\r\nstart with three questions to identify the broad shape:\r\n\r\n- **What is the video?** (one-sentence brief)\r\n- **How long?** (5-30s teaser \u002F 30-90s short \u002F 90s-3min explainer \u002F 3-10min film \u002F longer)\r\n- **What aspect ratio + target platform?** (1:1 \u002F 9:16 \u002F 16:9; X, IG, YouTube, internal, etc.)\r\n\r\nFrom the answer, classify the style category. The style determines which\r\nfollow-up questions to ask. **Do not ask all questions at once.** Ask 2-4 at a\r\ntime, listen, then proceed. Make reasonable assumptions whenever the user\r\nimplies an answer.\r\n\r\nFor complete intake patterns and per-style question banks, see\r\n**[references\u002Fintake.md](references\u002Fintake.md)**.\r\n\r\n### Step 2 — Brief\r\n\r\nOnce enough is known, produce a structured `brief.md` using the template in\r\n`assets\u002Fbrief.md.tmpl`. Stages:\r\n\r\n1. **Concept** — the one-sentence pitch + emotional north star\r\n2. **Scope** — duration, aspect, platform, deadline\r\n3. **Style** — visual references, brand constraints, tone\r\n4. **Scenes** — beat-by-beat breakdown (durations, content, target tool)\r\n5. **Audio** — narration \u002F music \u002F SFX \u002F silent (per scene if needed)\r\n6. **Deliverables** — file format, resolution, optional alternates (vertical cut, GIF, etc.)\r\n\r\nShow the brief to the user for confirmation before designing the team. **The\r\nbrief is the contract** — every downstream task references it.\r\n\r\n### Step 3 — Team design\r\n\r\nPick role archetypes from the library that fit this video. **Compose, don't\r\nclone.** Most videos need 4-7 profiles. The director is always present; the\r\nrest are picked by what the brief actually requires.\r\n\r\nFor the role library and per-style team compositions, see\r\n**[references\u002Frole-archetypes.md](references\u002Frole-archetypes.md)**.\r\n\r\nFor mapping role → which Hermes skills + toolsets it loads, see\r\n**[references\u002Ftool-matrix.md](references\u002Ftool-matrix.md)**.\r\n\r\n### Step 4 — Setup\r\n\r\nGenerate a setup script (`setup.sh`) and run it. The script:\r\n\r\n1. Creates the project workspace (`~\u002Fprojects\u002Fvideo-pipeline\u002F\u003Cslug>\u002F`)\r\n2. Copies any provided assets into `taste\u002F`, `audio\u002F`, `assets\u002F`\r\n3. Creates each Hermes profile via `hermes profile create --clone`\r\n4. Writes per-profile `SOUL.md` (personality + role definition)\r\n5. Configures profile YAML (toolsets, always_load skills, cwd)\r\n6. Writes `brief.md`, `TEAM.md`, and `taste\u002F` content\r\n7. Fires the initial `hermes kanban create` task assigned to the director\r\n\r\nUse `scripts\u002Fbootstrap_pipeline.py` to generate setup.sh from a brief +\r\nteam-design JSON. See **[references\u002Fkanban-setup.md](references\u002Fkanban-setup.md)**\r\nfor the setup script structure, profile config patterns, and the critical\r\n\"shared workspace\" rule.\r\n\r\n### Step 5 — Execute\r\n\r\nRun `setup.sh`. Then provide the user with monitoring commands:\r\n\r\n```bash\r\nhermes kanban watch --tenant \u003Cproject-tenant>     # live events\r\nhermes kanban list  --tenant \u003Cproject-tenant>     # board snapshot\r\nhermes dashboard                                   # visual board UI\r\n```\r\n\r\nThe director profile takes over from here, decomposing the work and routing\r\ntasks to specialist profiles via the kanban toolset.\r\n\r\n### Step 6 — Monitor and intervene\r\n\r\nStay engaged — the kanban runs autonomously but a stuck task or bad output\r\nneeds human (or AI) judgment.\r\n\r\nMonitoring patterns: poll `kanban list` periodically, inspect any RUNNING task\r\nthat exceeds its expected duration with `kanban show \u003Cid>`, and check\r\nheartbeats. When a worker's output fails review, the standard interventions are:\r\n\r\n1. Comment on the worker's task with specific feedback (`kanban_comment`)\r\n2. Create a re-run task with the original as parent\r\n3. Adjust the brief's scope and let the director re-decompose\r\n\r\nFor diagnostic patterns, intervention recipes, and the \"task is stuck\"\r\nplaybook, see **[references\u002Fmonitoring.md](references\u002Fmonitoring.md)**.\r\n\r\n## Reference: worked examples\r\n\r\nSix concrete pipelines covering very different video styles — narrative film,\r\nproduct\u002Fmarketing, music video, math\u002Falgorithm explainer, ASCII video, real-time\r\ninstallation — showing how the same workflow yields very different teams and\r\ntask graphs. See **[references\u002Fexamples.md](references\u002Fexamples.md)**.\r\n\r\n## Critical rules\r\n\r\n1. **Discovery before action.** Never start generating a brief or team without\r\n   asking at least the three baseline questions. A bad brief cascades through\r\n   the entire pipeline.\r\n\r\n2. **Match the team to the video.** Don't reuse the same 4-profile setup for\r\n   every job. A music video that doesn't have a beat-analysis profile will\r\n   misfire. A narrative film that doesn't have a writer profile will produce\r\n   incoherent scenes. See `references\u002Frole-archetypes.md`.\r\n\r\n3. **One workspace per project.** All profiles for a given video share the same\r\n   `dir:` workspace. Tasks pass artifacts via shared filesystem and structured\r\n   handoffs. **Every** `kanban_create` call passes\r\n   `workspace_kind=\"dir\"` + `workspace_path=\"\u003Cabsolute project path>\"`.\r\n\r\n4. **Tenant every project.** Use a project-specific tenant\r\n   (`--tenant \u003Cproject-slug>`). Keeps the dashboard scoped and prevents\r\n   cross-pollination with other ongoing kanbans.\r\n\r\n5. **Respect existing skills.** When a scene fits an existing skill, the\r\n   relevant renderer should load that skill via `--skill \u003Cname>` on its task\r\n   or `always_load` in its profile. Do not re-derive what a skill already\r\n   provides.\r\n\r\n6. **The director never executes.** Even with the full `kanban + terminal +\r\n   file` toolset, the director's `SOUL.md` rules forbid it from executing\r\n   work itself. It decomposes and routes only — every concrete task becomes\r\n   a `hermes kanban create` call to a specialist profile. The\r\n   `kanban-orchestrator` skill spells this out further.\r\n\r\n7. **Don't over-decompose.** A 30-second product video does NOT need 20 tasks.\r\n   Aim for the smallest task graph that still parallelizes well and exposes the\r\n   right human-review gates.\r\n\r\n8. **Verify API keys BEFORE firing.** External APIs (TTS, image-gen,\r\n   image-to-video) need keys in `${HERMES_HOME:-~\u002F.hermes}\u002F.env` or the user's secret store.\r\n   A worker that hits a missing-key error wastes a task slot. The setup\r\n   script's `check_key` helper aborts cleanly if a required key is missing.\r\n\r\n## File map\r\n\r\n```\r\nSKILL.md                            ← this file (workflow + rules)\r\nreferences\u002F\r\n  intake.md                         ← discovery question banks per style\r\n  role-archetypes.md                ← role library (writer, designer, animator, …)\r\n  tool-matrix.md                    ← skill + toolset mapping per role\r\n  kanban-setup.md                   ← setup script structure & profile config\r\n  monitoring.md                     ← watch + intervene patterns\r\n  examples.md                       ← six worked pipelines\r\nassets\u002F\r\n  brief.md.tmpl                     ← brief skeleton\r\n  setup.sh.tmpl                     ← setup script skeleton\r\n  soul.md.tmpl                      ← profile personality skeleton\r\nscripts\u002F\r\n  bootstrap_pipeline.py             ← generate setup.sh from brief + team JSON\r\n  monitor.py                        ← polling + intervention helpers\r\n```\r\n","# Kanban Video Orchestrator\r\n\r\nWrap any video request — from a 15-second product teaser to a 5-minute narrative\r\nshort to a music video to an ASCII loop — in a Hermes Kanban pipeline that\r\ndecomposes the work to specialized agent profiles.\r\n\r\nThis skill does **not** render anything itself. It is a meta-pipeline that:\r\n\r\n1. **Scopes** the request through targeted discovery\r\n2. **Designs** an appropriate team (which roles, which tools per role) based on the style\r\n3. **Generates** a setup script that creates Hermes profiles, project workspace, and the initial kanban task\r\n4. **Hands off** to the director profile, which decomposes via the kanban\r\n5. **Monitors** execution, helps intervene when tasks stall or fail\r\n\r\nThe actual rendering happens inside the kanban once it's running, via whichever\r\nexisting skills + tools fit the scenes — `ascii-video`, `manim-video`, `p5js`,\r\n`comfyui`, `touchdesigner-mcp`, `blender-mcp`, `songwriting-and-ai-music`,\r\n`heartmula`, external APIs, or plain Python with PIL + ffmpeg.\r\n\r\n## When NOT to use this skill\r\n\r\n- The video is one continuous procedural project that needs no specialists. Just write the code directly.\r\n- The user wants a quick one-shot conversion (e.g. \"convert this mp4 to a GIF\") — use ffmpeg directly.\r\n- The output is a static image, GIF, or audio-only artifact — use the matching specific skill (`ascii-art`, `gifs`, `meme-generation`, `songwriting-and-ai-music`).\r\n- The work fits a single existing skill cleanly (e.g. a pure ASCII video — just use `ascii-video`).\r\n\r\n## Workflow\r\n\r\n```\r\nDISCOVER  →  BRIEF  →  TEAM DESIGN  →  SETUP  →  EXECUTE  →  MONITOR\r\n```\r\n\r\n### Step 1 — Discover (ask the right questions)\r\n\r\nThe discovery process is **adaptive**: ask only what is actually needed. Always\r\nstart with three questions to identify the broad shape:\r\n\r\n- **What is the video?** (one-sentence brief)\r\n- **How long?** (5-30s teaser \u002F 30-90s short \u002F 90s-3min explainer \u002F 3-10min film \u002F longer)\r\n- **What aspect ratio + target platform?** (1:1 \u002F 9:16 \u002F 16:9; X, IG, YouTube, internal, etc.)\r\n\r\nFrom the answer, classify the style category. The style determines which\r\nfollow-up questions to ask. **Do not ask all questions at once.** Ask 2-4 at a\r\ntime, listen, then proceed. Make reasonable assumptions whenever the user\r\nimplies an answer.\r\n\r\nFor complete intake patterns and per-style question banks, see\r\n**[references\u002Fintake.md](references\u002Fintake.md)**.\r\n\r\n### Step 2 — Brief\r\n\r\nOnce enough is known, produce a structured `brief.md` using the template in\r\n`assets\u002Fbrief.md.tmpl`. Stages:\r\n\r\n1. **Concept** — the one-sentence pitch + emotional north star\r\n2. **Scope** — duration, aspect, platform, deadline\r\n3. **Style** — visual references, brand constraints, tone\r\n4. **Scenes** — beat-by-beat breakdown (durations, content, target tool)\r\n5. **Audio** — narration \u002F music \u002F SFX \u002F silent (per scene if needed)\r\n6. **Deliverables** — file format, resolution, optional alternates (vertical cut, GIF, etc.)\r\n\r\nShow the brief to the user for confirmation before designing the team. **The\r\nbrief is the contract** — every downstream task references it.\r\n\r\n### Step 3 — Team design\r\n\r\nPick role archetypes from the library that fit this video. **Compose, don't\r\nclone.** Most videos need 4-7 profiles. The director is always present; the\r\nrest are picked by what the brief actually requires.\r\n\r\nFor the role library and per-style team compositions, see\r\n**[references\u002Frole-archetypes.md](references\u002Frole-archetypes.md)**.\r\n\r\nFor mapping role → which Hermes skills + toolsets it loads, see\r\n**[references\u002Ftool-matrix.md](references\u002Ftool-matrix.md)**.\r\n\r\n### Step 4 — Setup\r\n\r\nGenerate a setup script (`setup.sh`) and run it. The script:\r\n\r\n1. Creates the project workspace (`~\u002Fprojects\u002Fvideo-pipeline\u002F\u003Cslug>\u002F`)\r\n2. Copies any provided assets into `taste\u002F`, `audio\u002F`, `assets\u002F`\r\n3. Creates each Hermes profile via `hermes profile create --clone`\r\n4. Writes per-profile `SOUL.md` (personality + role definition)\r\n5. Configures profile YAML (toolsets, always_load skills, cwd)\r\n6. Writes `brief.md`, `TEAM.md`, and `taste\u002F` content\r\n7. Fires the initial `hermes kanban create` task assigned to the director\r\n\r\nUse `scripts\u002Fbootstrap_pipeline.py` to generate setup.sh from a brief +\r\nteam-design JSON. See **[references\u002Fkanban-setup.md](references\u002Fkanban-setup.md)**\r\nfor the setup script structure, profile config patterns, and the critical\r\n\"shared workspace\" rule.\r\n\r\n### Step 5 — Execute\r\n\r\nRun `setup.sh`. Then provide the user with monitoring commands:\r\n\r\n```bash\r\nhermes kanban watch --tenant \u003Cproject-tenant>     # live events\r\nhermes kanban list  --tenant \u003Cproject-tenant>     # board snapshot\r\nhermes dashboard                                   # visual board UI\r\n```\r\n\r\nThe director profile takes over from here, decomposing the work and routing\r\ntasks to specialist profiles via the kanban toolset.\r\n\r\n### Step 6 — Monitor and intervene\r\n\r\nStay engaged — the kanban runs autonomously but a stuck task or bad output\r\nneeds human (or AI) judgment.\r\n\r\nMonitoring patterns: poll `kanban list` periodically, inspect any RUNNING task\r\nthat exceeds its expected duration with `kanban show \u003Cid>`, and check\r\nheartbeats. When a worker's output fails review, the standard interventions are:\r\n\r\n1. Comment on the worker's task with specific feedback (`kanban_comment`)\r\n2. Create a re-run task with the original as parent\r\n3. Adjust the brief's scope and let the director re-decompose\r\n\r\nFor diagnostic patterns, intervention recipes, and the \"task is stuck\"\r\nplaybook, see **[references\u002Fmonitoring.md](references\u002Fmonitoring.md)**.\r\n\r\n## Reference: worked examples\r\n\r\nSix concrete pipelines covering very different video styles — narrative film,\r\nproduct\u002Fmarketing, music video, math\u002Falgorithm explainer, ASCII video, real-time\r\ninstallation — showing how the same workflow yields very different teams and\r\ntask graphs. See **[references\u002Fexamples.md](references\u002Fexamples.md)**.\r\n\r\n## Critical rules\r\n\r\n1. **Discovery before action.** Never start generating a brief or team without\r\n   asking at least the three baseline questions. A bad brief cascades through\r\n   the entire pipeline.\r\n\r\n2. **Match the team to the video.** Don't reuse the same 4-profile setup for\r\n   every job. A music video that doesn't have a beat-analysis profile will\r\n   misfire. A narrative film that doesn't have a writer profile will produce\r\n   incoherent scenes. See `references\u002Frole-archetypes.md`.\r\n\r\n3. **One workspace per project.** All profiles for a given video share the same\r\n   `dir:` workspace. Tasks pass artifacts via shared filesystem and structured\r\n   handoffs. **Every** `kanban_create` call passes\r\n   `workspace_kind=\"dir\"` + `workspace_path=\"\u003Cabsolute project path>\"`.\r\n\r\n4. **Tenant every project.** Use a project-specific tenant\r\n   (`--tenant \u003Cproject-slug>`). Keeps the dashboard scoped and prevents\r\n   cross-pollination with other ongoing kanbans.\r\n\r\n5. **Respect existing skills.** When a scene fits an existing skill, the\r\n   relevant renderer should load that skill via `--skill \u003Cname>` on its task\r\n   or `always_load` in its profile. Do not re-derive what a skill already\r\n   provides.\r\n\r\n6. **The director never executes.** Even with the full `kanban + terminal +\r\n   file` toolset, the director's `SOUL.md` rules forbid it from executing\r\n   work itself. It decomposes and routes only — every concrete task becomes\r\n   a `hermes kanban create` call to a specialist profile. The\r\n   `kanban-orchestrator` skill spells this out further.\r\n\r\n7. **Don't over-decompose.** A 30-second product video does NOT need 20 tasks.\r\n   Aim for the smallest task graph that still parallelizes well and exposes the\r\n   right human-review gates.\r\n\r\n8. **Verify API keys BEFORE firing.** External APIs (TTS, image-gen,\r\n   image-to-video) need keys in `${HERMES_HOME:-~\u002F.hermes}\u002F.env` or the user's secret store.\r\n   A worker that hits a missing-key error wastes a task slot. The setup\r\n   script's `check_key` helper aborts cleanly if a required key is missing.\r\n\r\n## File map\r\n\r\n```\r\nSKILL.md                            ← this file (workflow + rules)\r\nreferences\u002F\r\n  intake.md                         ← discovery question banks per style\r\n  role-archetypes.md                ← role library (writer, designer, animator, …)\r\n  tool-matrix.md                    ← skill + toolset mapping per role\r\n  kanban-setup.md                   ← setup script structure & profile config\r\n  monitoring.md                     ← watch + intervene patterns\r\n  examples.md                       ← six worked pipelines\r\nassets\u002F\r\n  brief.md.tmpl                     ← brief skeleton\r\n  setup.sh.tmpl                     ← setup script skeleton\r\n  soul.md.tmpl                      ← profile personality skeleton\r\nscripts\u002F\r\n  bootstrap_pipeline.py             ← generate setup.sh from brief + team JSON\r\n  monitor.py                        ← polling + intervention helpers\r\n```\r\n",{"name":5,"author":60,"license":15,"version":63,"metadata":64,"platforms":96,"description":7},[61,62],"SHL0MS","alt-glitch","1.0.0",{"hermes":65},{"tags":66,"credits":71,"related_skills":72},[67,37,68,69,70],"video","multi-agent","orchestration","production-pipeline","The single-project workspace layout, profile-config patching pattern,\nSOUL.md-per-profile model, TEAM.md task-graph convention, and\n`--workspace dir:\u003Cpath>` discipline are adapted from alt-glitch's\noriginal multi-agent video pipeline at\nhttps:\u002F\u002Fgithub.com\u002FNousResearch\u002Fkanban-video-pipeline.",[73,74,75,76,77,78,79,80,40,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95],"kanban-orchestrator","kanban-worker","ascii-video","manim-video","p5js","comfyui","touchdesigner-mcp","blender-mcp","ascii-art","songwriting-and-ai-music","heartmula","songsee","spotify","youtube-content","claude-design","excalidraw","architecture-diagram","concept-diagrams","baoyu-comic","baoyu-infographic","humanizer","gif-search","meme-generation",[97,98,99],"linux","macos","windows","main","0d1434f89f4bc6b07e9d272accd8c1013768820f","archive\u002Fskill-md-reviewed\u002Fsha256\u002Fbe\u002Fbe54d5f681fff01e2ce3b4684d9644382343f8e12a212c6ca482b90a654c41f6",0,"https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills","https:\u002F\u002Fgithub.com\u002Fzapabob","The Open University of JAPAN　2019-now\r\n\r\nTokyo University of Science　2007-2010","2026-10-26 16:29:01",{"source":8,"skillName":5,"commands":109,"prompt":119,"zipAvailable":120},[110,113,116],{"tool":111,"command":112},"npx","npx skills add zapabob\u002FSkills --skill kanban-video-orchestrator",{"tool":114,"command":115},"bunx","bunx skills add zapabob\u002FSkills --skill kanban-video-orchestrator",{"tool":117,"command":118},"pnpm","pnpm dlx skills add zapabob\u002FSkills --skill kanban-video-orchestrator","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Farchive\u002Fskill-md-reviewed\u002Fsha256\u002Fbe\u002Fbe54d5f681fff01e2ce3b4684d9644382343f8e12a212c6ca482b90a654c41f6\u002FSKILL.md",true,[],[123,151,183,207,233,250],{"skillNo":124,"name":125,"nameZh":6,"description":126,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":127,"categories":128,"keywords":136,"publishedAt":55},"01M6GJTE1GESGGF9K028Y7753E","xurl","X\u002FTwitter via xurl CLI: post, search, DM, media, v2 API.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fxurl\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[129,132,133],{"categoryNo":130,"categoryName":131,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000010","后端与接口",{"categoryNo":27,"categoryName":28,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":134,"categoryName":135,"taxonomyType":32,"taxonomyTypeDesc":33},"01JBSEEDCAT000000000000030","开发工程师",[125,137,98,138,139,97,99,140,51,141,142,143,144,145,146,147,100,148,149,150],"openclaw","cli","github","official","https","social-media","api","Media","hermes","hermes agent","twitter","social","via","post",{"skillNo":152,"name":153,"nameZh":6,"description":154,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":155,"categories":156,"keywords":166,"publishedAt":55},"01M6GJT8S1PXQ8WPXT1HSNBWH2","writing-plans","Write implementation plans: bite-sized tasks, paths, code.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwriting-plans\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[157,158,161,162,163],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":159,"categoryName":160,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000018","写作与文案",{"categoryNo":27,"categoryName":28,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":134,"categoryName":135,"taxonomyType":32,"taxonomyTypeDesc":33},{"categoryNo":164,"categoryName":165,"taxonomyType":32,"taxonomyTypeDesc":33},"01JBSEEDCAT000000000000036","内容创作者",[167,168,169,46,98,170,97,171,99,172,173,174,175,176,177,178,179,180,181,182],"writing","writing plans","plans","review","workflow","documentation","tasks","development","code","planning","code-review","subagent","write","superpowers","paths","test-driven-development",{"skillNo":184,"name":185,"nameZh":6,"description":186,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":187,"categories":188,"keywords":190,"publishedAt":55},"01M6GJST0H5X7WHBJ3BZK0DVKH","whisper","OpenAI's general-purpose speech recognition model. Supports 99 languages, transcription, translation to English, and language identification. Six model sizes from tiny (39M params) to large (1550M params). Use for speech-to-text, podcast transcription, or multilingual audio processing. Best for robust, multilingual ASR.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwhisper\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[189],{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},[185,98,97,191,192,193,194,195,49,196,197,198,199,200,201,202,203,204,205,206],"openai","speech","multimodal","Transformers","ASR","text","model","speech recognition","use","general","multilingual","translation","english","identification","processing","Audio Processing",{"skillNo":208,"name":209,"nameZh":6,"description":210,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":211,"categories":212,"keywords":216,"publishedAt":55},"01M6GJSMRQ8EETZTYMG72P70EA","weights-and-biases","W&B: log ML experiments, sweeps, model registry, dashboards.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fweights-and-biases\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[213],{"categoryNo":214,"categoryName":215,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000013","分析与可视化",[217,98,218,219,99,220,221,222,223,224,225,197,226,227,44,228,229,230,231,232],"weights and biases","tensorflow","experiments","pytorch","visualization","registry","real-time","tuning","dashboards","experiment","log","Experiment Tracking","Model Registry","Hyperparameter Tuning","WandB","Real-Time Visualization",{"skillNo":234,"name":235,"nameZh":6,"description":236,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":237,"categories":238,"keywords":240,"publishedAt":55},"01M6GJS19KDPRY112JTR326S06","watchers","Poll RSS, JSON APIs, and GitHub with watermark dedup.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fwatchers\u002Fvariants\u002Fhermes-agent-optional\u002FSKILL.md",[239],{"categoryNo":27,"categoryName":28,"taxonomyType":21,"taxonomyTypeDesc":22},[235,98,241,139,97,242,243,244,245,246,247,145,146,248,249],"terminal","automation","http","monitoring","cron","json","rss","devops","with",{"skillNo":251,"name":252,"nameZh":6,"description":253,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":254,"categories":255,"keywords":258,"publishedAt":55},"01M6GJRHJD2A6Q830SWF9W48BC","serving-llms-vllm","vLLM: high-throughput LLM serving, OpenAI API, quantization.","https:\u002F\u002Fgithub.com\u002Fzapabob\u002FSkills\u002Fblob\u002Fmain\u002Fregistry\u002Fskills\u002Fvllm\u002Fvariants\u002Fhermes-agent\u002FSKILL.md",[256,257],{"categoryNo":130,"categoryName":131,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":134,"categoryName":135,"taxonomyType":32,"taxonomyTypeDesc":33},[259,260,261,97,262,191,263,264,265,194,266,143,267,268,269,270,271,272,273,274],"llms","vLLM","serving llms vllm","research","openai api","tensor","Inference Serving","Production","llm serving","High Throughput","PagedAttention","Continuous Batching","Quantization","Tensor Parallelism","high-throughput","throughput"]