[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6BY19SJD7SQQX00W2674RFB":3,"skill-same:01M6BY19SJD7SQQX00W2674RFB":68},{"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":40,"linkOnly":41,"markdownContent":42,"markdownBody":43,"frontMatter":44,"refName":45,"commitSha":46,"skillDirectory":47,"forksCount":48,"repositoryHtmlUrl":49,"repositoryHomepageUrl":50,"ownerHtmlUrl":51,"ownerBlogUrl":12,"ownerDescription":52,"updatedAt":53,"install":54,"resources":67},"01M6BY19SJD7SQQX00W2674RFB","native-game-performance",null,"Make a native NodeTool game run at the display rate: frame and tick targets, budgets per game type, the authoring rules that set engine cost, and the measure-fix loop with nodetool game bench, capture and perf. Load before building a 3D game, at every build milestone, and when a game stutters, drops frames or runs slowly.","nodetool-ai\u002Fnodetool","nodetool-ai","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F164873569?v=4","01M6BXPFDW36K9K1MR8ZH63HG2","nodetool.ai","01M6BXPFJ893XM219GQN60B0WK",635,"AGPL-3.0","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fnative-game-performance\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39],"native","game","native game performance","native game","performance","nodetool","when","rate","at","display","make","authoring","milestone","fix","before building","engine","loop","3D","SET","run","perf","2026-11-10 10:12:30",false,"---\nname: native-game-performance\ndescription: \"Make a native NodeTool game run at the display rate: frame and tick targets, budgets per game type, the authoring rules that set engine cost, and the measure-fix loop with nodetool game bench, capture and perf. Load before building a 3D game, at every build milestone, and when a game stutters, drops frames or runs slowly.\"\n---\n\n# Native game performance\n\nA game is done when it holds its frame target on the slowest machine it ships\nto. Measure at every milestone. A number from a tool decides, not a look at the\neditor.\n\n## Targets\n\n- The simulation ticks at 60 Hz. Tick p95 stays at or under 4 ms\n  (`performance.budgets.tickMs`, default 4). That leaves the rest of each\n  16.6 ms frame to rendering.\n- The player renders at the display rate. On the slowest target Mac, frame p95\n  stays within 16.6 ms for a 60 Hz display and within 8.3 ms for a 120 Hz\n  display (`performance.budgets.frameMs`, default 16.6). Set `frameMs: 8.3`\n  when the game targets ProMotion displays.\n- Shader programs stay flat during play. Growth means a material switches its\n  program variant every frame, which costs more than any draw.\n\nSet the time budgets with the other budgets:\n`edit_native_game {game_id, ops: [{op: \"set_performance\", performance: {budgets: {tickMs: 4, frameMs: 8.3, drawCalls: 400, triangles: 2000000}}}]}`.\n`set_performance` replaces the whole `performance` object, so send\n`cullLayers` again when the game has them.\n\n## Budgets per game type\n\nStarting budgets for the slowest target Mac. Draw calls and triangles count\nevery pass: the scene, each shadow cascade and the soft-particle depth pass.\n\n| Budget | Arena | Platformer | Open 3D world | Flight or driving |\n|---|---|---|---|---|\n| Draw calls per frame | 300 | 300 | 500 | 500 |\n| Triangles per frame | 500k | 500k | 2M | 3M |\n| Shadow cascades | none, one map with `extent` | 2 | 3 | 3 |\n| Cascade `maxDistance` | — | 60 to 120 | 300 to 800 | 1000 to 2000 |\n| `shadows.mapSize` | 1024 | 1024 | 2048 | 2048 |\n| Live particles | 1000 | 1000 | 2000 | 4000 |\n| Soft emitters (`softness > 0`) | 1 | 2 | 4 | 4 |\n| Collider triangles (all trimeshes) | 5k | 20k | 50k | 20k |\n| Script behaviors | 1 to 5 | 1 to 5 | 10 | 1 to 3 |\n| Active entities in a scripted scene | 300 | 300 | 1000 | 300 |\n\nRIDGELINE, the shipped flight showcase, measures 380 draw calls, 2.9M\ntriangles, about 4000 particles, a 13.5k-triangle collider and 3 scripts.\n\n## Authoring rules\n\nEach rule removes work the engine repeats every frame or every tick.\n\n- **`castShadow: false` on terrain and ground.** Every casting mesh draws again\n  into each cascade, and ground still receives shadows. In a scene with\n  cascades, validation warns with `large_shadow_caster` on a casting mesh 100\n  units or wider.\n- **2 or 3 cascades with a real `maxDistance`.** Each cascade is another pass\n  over every caster inside it. A `maxDistance` far beyond what the player sees\n  spreads the map over kilometres and makes nearly everything a caster.\n- **Merge repeated props.** The engine does not instance: each model entity is\n  its own draw calls and its own cloned materials. Bake trees, rocks and fence\n  posts into a few merged models.\n- **Split large meshes into chunks** of a few hundred metres. The renderer culls\n  whole meshes, so one huge terrain mesh always draws in full, in every pass.\n- **Use a reduced collision mesh.** The physics world is copied and restored\n  every tick to keep replays exact, and that copy grows with collider triangles.\n  Build a decimated `triangleMesh` collider, never the render mesh.\n- **Use one controller script.** Each script call pays realm setup and JSON\n  transport. One script that drives many entities costs less than one script\n  per entity. Read the world with `world.get` and `world.query` with `near` and\n  `radius`. Reading `input.world` serializes every active entity (1 MiB limit).\n- **Send `setProp` and `ui` commands only when the value changes.** Props travel\n  in the script input and the snapshot. A changed HUD value repaints the whole\n  HUD canvas and uploads it as a texture. Round a speed or altitude readout so\n  it changes a few times a second.\n- **Keep the sun still under a procedural sky.** The sky cube and its\n  environment map rebuild whenever the sun direction changes. A day cycle that\n  turns the sun every tick rebuilds them every tick.\n- **`presentation.hudWidth` is the render resolution.** The exported player\n  sizes the 3D canvas to `hudWidth` by `hudWidth \u002F aspectRatio`. Every\n  full-screen pass scales with it. 1280 is enough for most games.\n- **Bloom costs a mip chain.** Bloom adds a bright pass and five blur levels at\n  full resolution. Keep `threshold` high so only real highlights bloom, and\n  leave vignette, `fxaa` and `smaa` off unless the look needs them.\n- **Keep soft particles tight.** `softness` makes the renderer draw scene depth\n  once more for every geometry near any soft emitter. Set it only on emitters\n  that touch geometry, and keep those emitters small.\n\n## The measure-fix loop\n\nRun this loop at every milestone: after the blockout, after the art pass, after\nthe effects pass, and before you report the game as done.\n\n1. Record a route. A route is a JSON array of input frames, one per tick:\n   `{\"pressed\": [], \"justPressed\": [], \"axes\": {\"throttle\": 1}, \"look\": {\"x\": 0, \"y\": 0}}`.\n   It must reach the busiest moment of play: the most enemies, particles and\n   scenery in view. An idle start screen measures nothing.\n2. Measure the tick:\n   `nodetool game bench game.json --inputs route.json --warmup 600 --ticks 600 --fail-on-budget --json`.\n   It reports `tickMs`, `scriptMs`, time per system, physics counts and sampled\n   allocation bytes per tick.\n3. Measure the busiest frame's counters:\n   `nodetool game capture game.json --ticks \u003Cbusiest tick> --inputs route.json --out frame.png --fail-on-budget --json`.\n   Read `stats.drawCalls`, `stats.triangles`, `stats.programs` and\n   `budget.overruns`. `stats.renderMs` is CPU submission time only.\n4. Measure real play on the GPU:\n   `nodetool game perf game.json --inputs route.json --seconds 10 --fail-on-budget --profile play.cpuprofile`.\n   It exports the game, replays the route, resumes live play and reports frame\n   time p50\u002Fp95\u002Fp99, fps, draw calls, triangles and shader program growth.\n5. Fix the largest cost first, then measure again with the same route. Stop when\n   all three commands exit 0.\n\nA 2D game runs step 2 only: `game capture --fail-on-budget` and `game perf`\nmeasure 3D games. Each command takes a game document JSON or a shipped example\nfile, and\n`--assets-dir` for installed media named `\u003Cfull-asset-id>.\u003Cextension>`.\n`game perf` uses uncapped frames, so its frame time is the real cost of a frame\non this machine. `--software` runs without a GPU, and its times say nothing\nabout hardware.\n\n### Reading the counters\n\n- Tick over budget: compare `perSystemMs`. High `scripts` points to script\n  count, `input.world` reads or props. High `character` or `physics` points to\n  collider size.\n- Draw calls over budget: count casting meshes and cascades first, then\n  repeated props.\n- Triangles over budget: shadow passes multiply triangles. Turn off\n  `castShadow` on large meshes and lower the cascade count before you simplify\n  models.\n- Frame time over budget with counters in budget: halve `hudWidth` and measure\n  again. A big drop means the frame is fill-bound (resolution, bloom, soft\n  particles). No drop means the frame is CPU-bound.\n- Program growth above 0: a material or light changes every frame. Look for\n  lights added, removed or toggled during play, and materials swapped per frame.\n\n### Reading a profile\n\n`game perf --profile` lists the functions with the most self time and flags\nsuspects. In steady play, renderer time belongs in `renderBufferDirect`,\n`setProgram` and `projectObject`. Each of these in steady play is a bug:\n`getProgram` or `getParameters` (a program rebuilt each frame), `traverse` with\na large share (a scene walk each frame), `takeSnapshot`, `JSON.stringify` or\n`structuredClone` (a serializer in the frame or tick). Report it with the\nprofile path, then fix the authoring cause or the engine.\n\n## Without a shell\n\nInside NodeTool there is no tool that times frames. Apply the budgets and\nauthoring rules, set them with `set_performance`, and check counts by reading\nthe document with `get_native_game {game_id, view: \"full\"}`. Give the user the\n`nodetool game perf` command for the game, and when they ask for an export,\nbuild it with `build_native_game`.\n\n## Editor numbers\n\nEditor play runs slower than the exported player. It inspects the session every\ntick, copies state into the editor panels, and draws selection outlines and\ndebug views. Judge performance on the export with `game perf`, never on editor\nfps.\n","# Native game performance\n\nA game is done when it holds its frame target on the slowest machine it ships\nto. Measure at every milestone. A number from a tool decides, not a look at the\neditor.\n\n## Targets\n\n- The simulation ticks at 60 Hz. Tick p95 stays at or under 4 ms\n  (`performance.budgets.tickMs`, default 4). That leaves the rest of each\n  16.6 ms frame to rendering.\n- The player renders at the display rate. On the slowest target Mac, frame p95\n  stays within 16.6 ms for a 60 Hz display and within 8.3 ms for a 120 Hz\n  display (`performance.budgets.frameMs`, default 16.6). Set `frameMs: 8.3`\n  when the game targets ProMotion displays.\n- Shader programs stay flat during play. Growth means a material switches its\n  program variant every frame, which costs more than any draw.\n\nSet the time budgets with the other budgets:\n`edit_native_game {game_id, ops: [{op: \"set_performance\", performance: {budgets: {tickMs: 4, frameMs: 8.3, drawCalls: 400, triangles: 2000000}}}]}`.\n`set_performance` replaces the whole `performance` object, so send\n`cullLayers` again when the game has them.\n\n## Budgets per game type\n\nStarting budgets for the slowest target Mac. Draw calls and triangles count\nevery pass: the scene, each shadow cascade and the soft-particle depth pass.\n\n| Budget | Arena | Platformer | Open 3D world | Flight or driving |\n|---|---|---|---|---|\n| Draw calls per frame | 300 | 300 | 500 | 500 |\n| Triangles per frame | 500k | 500k | 2M | 3M |\n| Shadow cascades | none, one map with `extent` | 2 | 3 | 3 |\n| Cascade `maxDistance` | — | 60 to 120 | 300 to 800 | 1000 to 2000 |\n| `shadows.mapSize` | 1024 | 1024 | 2048 | 2048 |\n| Live particles | 1000 | 1000 | 2000 | 4000 |\n| Soft emitters (`softness > 0`) | 1 | 2 | 4 | 4 |\n| Collider triangles (all trimeshes) | 5k | 20k | 50k | 20k |\n| Script behaviors | 1 to 5 | 1 to 5 | 10 | 1 to 3 |\n| Active entities in a scripted scene | 300 | 300 | 1000 | 300 |\n\nRIDGELINE, the shipped flight showcase, measures 380 draw calls, 2.9M\ntriangles, about 4000 particles, a 13.5k-triangle collider and 3 scripts.\n\n## Authoring rules\n\nEach rule removes work the engine repeats every frame or every tick.\n\n- **`castShadow: false` on terrain and ground.** Every casting mesh draws again\n  into each cascade, and ground still receives shadows. In a scene with\n  cascades, validation warns with `large_shadow_caster` on a casting mesh 100\n  units or wider.\n- **2 or 3 cascades with a real `maxDistance`.** Each cascade is another pass\n  over every caster inside it. A `maxDistance` far beyond what the player sees\n  spreads the map over kilometres and makes nearly everything a caster.\n- **Merge repeated props.** The engine does not instance: each model entity is\n  its own draw calls and its own cloned materials. Bake trees, rocks and fence\n  posts into a few merged models.\n- **Split large meshes into chunks** of a few hundred metres. The renderer culls\n  whole meshes, so one huge terrain mesh always draws in full, in every pass.\n- **Use a reduced collision mesh.** The physics world is copied and restored\n  every tick to keep replays exact, and that copy grows with collider triangles.\n  Build a decimated `triangleMesh` collider, never the render mesh.\n- **Use one controller script.** Each script call pays realm setup and JSON\n  transport. One script that drives many entities costs less than one script\n  per entity. Read the world with `world.get` and `world.query` with `near` and\n  `radius`. Reading `input.world` serializes every active entity (1 MiB limit).\n- **Send `setProp` and `ui` commands only when the value changes.** Props travel\n  in the script input and the snapshot. A changed HUD value repaints the whole\n  HUD canvas and uploads it as a texture. Round a speed or altitude readout so\n  it changes a few times a second.\n- **Keep the sun still under a procedural sky.** The sky cube and its\n  environment map rebuild whenever the sun direction changes. A day cycle that\n  turns the sun every tick rebuilds them every tick.\n- **`presentation.hudWidth` is the render resolution.** The exported player\n  sizes the 3D canvas to `hudWidth` by `hudWidth \u002F aspectRatio`. Every\n  full-screen pass scales with it. 1280 is enough for most games.\n- **Bloom costs a mip chain.** Bloom adds a bright pass and five blur levels at\n  full resolution. Keep `threshold` high so only real highlights bloom, and\n  leave vignette, `fxaa` and `smaa` off unless the look needs them.\n- **Keep soft particles tight.** `softness` makes the renderer draw scene depth\n  once more for every geometry near any soft emitter. Set it only on emitters\n  that touch geometry, and keep those emitters small.\n\n## The measure-fix loop\n\nRun this loop at every milestone: after the blockout, after the art pass, after\nthe effects pass, and before you report the game as done.\n\n1. Record a route. A route is a JSON array of input frames, one per tick:\n   `{\"pressed\": [], \"justPressed\": [], \"axes\": {\"throttle\": 1}, \"look\": {\"x\": 0, \"y\": 0}}`.\n   It must reach the busiest moment of play: the most enemies, particles and\n   scenery in view. An idle start screen measures nothing.\n2. Measure the tick:\n   `nodetool game bench game.json --inputs route.json --warmup 600 --ticks 600 --fail-on-budget --json`.\n   It reports `tickMs`, `scriptMs`, time per system, physics counts and sampled\n   allocation bytes per tick.\n3. Measure the busiest frame's counters:\n   `nodetool game capture game.json --ticks \u003Cbusiest tick> --inputs route.json --out frame.png --fail-on-budget --json`.\n   Read `stats.drawCalls`, `stats.triangles`, `stats.programs` and\n   `budget.overruns`. `stats.renderMs` is CPU submission time only.\n4. Measure real play on the GPU:\n   `nodetool game perf game.json --inputs route.json --seconds 10 --fail-on-budget --profile play.cpuprofile`.\n   It exports the game, replays the route, resumes live play and reports frame\n   time p50\u002Fp95\u002Fp99, fps, draw calls, triangles and shader program growth.\n5. Fix the largest cost first, then measure again with the same route. Stop when\n   all three commands exit 0.\n\nA 2D game runs step 2 only: `game capture --fail-on-budget` and `game perf`\nmeasure 3D games. Each command takes a game document JSON or a shipped example\nfile, and\n`--assets-dir` for installed media named `\u003Cfull-asset-id>.\u003Cextension>`.\n`game perf` uses uncapped frames, so its frame time is the real cost of a frame\non this machine. `--software` runs without a GPU, and its times say nothing\nabout hardware.\n\n### Reading the counters\n\n- Tick over budget: compare `perSystemMs`. High `scripts` points to script\n  count, `input.world` reads or props. High `character` or `physics` points to\n  collider size.\n- Draw calls over budget: count casting meshes and cascades first, then\n  repeated props.\n- Triangles over budget: shadow passes multiply triangles. Turn off\n  `castShadow` on large meshes and lower the cascade count before you simplify\n  models.\n- Frame time over budget with counters in budget: halve `hudWidth` and measure\n  again. A big drop means the frame is fill-bound (resolution, bloom, soft\n  particles). No drop means the frame is CPU-bound.\n- Program growth above 0: a material or light changes every frame. Look for\n  lights added, removed or toggled during play, and materials swapped per frame.\n\n### Reading a profile\n\n`game perf --profile` lists the functions with the most self time and flags\nsuspects. In steady play, renderer time belongs in `renderBufferDirect`,\n`setProgram` and `projectObject`. Each of these in steady play is a bug:\n`getProgram` or `getParameters` (a program rebuilt each frame), `traverse` with\na large share (a scene walk each frame), `takeSnapshot`, `JSON.stringify` or\n`structuredClone` (a serializer in the frame or tick). Report it with the\nprofile path, then fix the authoring cause or the engine.\n\n## Without a shell\n\nInside NodeTool there is no tool that times frames. Apply the budgets and\nauthoring rules, set them with `set_performance`, and check counts by reading\nthe document with `get_native_game {game_id, view: \"full\"}`. Give the user the\n`nodetool game perf` command for the game, and when they ask for an export,\nbuild it with `build_native_game`.\n\n## Editor numbers\n\nEditor play runs slower than the exported player. It inspects the session every\ntick, copies state into the editor panels, and draws selection outlines and\ndebug views. Judge performance on the export with `game perf`, never on editor\nfps.\n",{"name":5,"description":7},"main","c7226baa82daca42a19a5855c116544f97aa8bac","packages\u002Fsystem-skills\u002Fnative-game-performance",91,"https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool","https:\u002F\u002Fnodetool.ai\u002F","https:\u002F\u002Fgithub.com\u002Fnodetool-ai","No-code AI development","2026-11-15 03:27:49",{"source":8,"skillName":5,"commands":55,"prompt":65,"zipAvailable":66},[56,59,62],{"tool":57,"command":58},"npx","npx skills add nodetool-ai\u002Fnodetool --skill native-game-performance",{"tool":60,"command":61},"bunx","bunx skills add nodetool-ai\u002Fnodetool --skill native-game-performance",{"tool":63,"command":64},"pnpm","pnpm dlx skills add nodetool-ai\u002Fnodetool --skill native-game-performance","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fnative-game-performance\u002FSKILL.md",true,[],[69,78,113,143,158,174],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":77},"01M742MX3Y2ZYK884ZXTRFKB9P","humanizer","Rewrite AI-sounding text so it reads like the writer without changing what it says.\nUse when editing or reviewing prose for AI tells: not-X-but-Y contrasts, one-line\nclosers, staged openers, forced triads, dashes everywhere, inflated claims, sales\nlanguage, stock AI words, bold labels, or filler.","MIT","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fhumanizer\u002FSKILL.md",[],[],"2026-11-11 15:11:14",{"skillNo":79,"name":24,"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":93,"publishedAt":40},"01M6BY4D93R0Q9400ZHJ04DXP1","Use the NodeTool MCP server to build, validate, run, and debug visual AI workflows, and to work with assets, collections, and jobs. Use when the user mentions NodeTool or asks for a workflow graph, a node, or a media generation run on their NodeTool server.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fplugins\u002Fchatgpt\u002Fnodetool\u002Fskills\u002Fnodetool\u002FSKILL.md",[83,88],{"categoryNo":84,"categoryName":85,"taxonomyType":86,"taxonomyTypeDesc":87},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":89,"categoryName":90,"taxonomyType":91,"taxonomyTypeDesc":92},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[24,94,95,96,25,97,98,99,100,101,102,103,104,105,106,107,108,109,38,110,111,112],"MCP","Media","workflows","generation","media generation","with","visual","use","collections","for","node","work","debug","graph","validate","assets","server","mcp server","workflow",{"skillNo":114,"name":115,"nameZh":6,"description":116,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":117,"categories":118,"keywords":122,"publishedAt":40},"01M6BY4AEB05ZPBRVD078ATR2H","wan-2-6-prompting","Prompt Alibaba's Wan 2.6 video — text-to-video with a global style line plus timing-bracketed shots, and image-to-video that describes only what changes over time. Covers the shot-mode switch per provider, the prompt budget, negative prompts, motion intensity wording, audio input and prompt expansion. Use whenever the model id names a Wan 2.6 video route (alibaba\u002Fwan-2.6\u002Ftext-to-video, alibaba\u002Fwan-2.6\u002Fimage-to-video, wan-video\u002Fwan-2.6-t2v, wan-video\u002Fwan-2.6-i2v, wan-video\u002Fwan2.6-i2v-flash) on generate_video, animate_image or a TextToVideo node. Not for Wan 2.5 or 2.7, or the Wan 2.6 image routes.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fwan-2-6-prompting\u002FSKILL.md",[119],{"categoryNo":120,"categoryName":121,"taxonomyType":86,"taxonomyTypeDesc":87},"01JBSEEDCAT000000000000016","图像与多媒体",[123,124,125,126,127,128,129,104,130,131,132,133,134,135,136,137,138,139,140,141,142],"prompting","wan 2 6 prompting","line","expansion","input","image-to-video","text-to-video","route","only","mode","time","flash","Budget","generate","text","image","animate","model","Audio","Video",{"skillNo":144,"name":145,"nameZh":6,"description":146,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":147,"categories":148,"keywords":151,"publishedAt":40},"01M6BY47T4748ZG5ZMTAR97ZWZ","video-workflow","Build a reusable NodeTool video workflow for batches or repeated inputs. For a single finished piece, use a storyboard skill.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fvideo-workflow\u002FSKILL.md",[149,150],{"categoryNo":120,"categoryName":121,"taxonomyType":86,"taxonomyTypeDesc":87},{"categoryNo":84,"categoryName":85,"taxonomyType":86,"taxonomyTypeDesc":87},[112,152,142,153,154,155,156,101,157,24,103],"video workflow","reusable","storyboard","build","batches","single",{"skillNo":159,"name":160,"nameZh":6,"description":161,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":162,"categories":163,"keywords":165,"publishedAt":40},"01M6BY4587N2SYN315SXGEMZC3","video-clone","Analyze a reference video and recreate its shot structure with a new product or cast in NodeTool.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fvideo-clone\u002FSKILL.md",[164],{"categoryNo":120,"categoryName":121,"taxonomyType":86,"taxonomyTypeDesc":87},[166,167,142,99,168,169,24,170,171,172,173],"clone","video clone","cast","new","analyze","product","structure","reference",{"skillNo":175,"name":176,"nameZh":6,"description":177,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":178,"categories":179,"keywords":181,"publishedAt":40},"01M6BY4257C2KQYYMD9EK6DK8E","video-audio-continuity","Keep sound continuous across a multi-scene piece cut from generated video — why one clip per scene hard-cuts the audio at every boundary, when to write all the scenes into a single generation instead, the sound-design vocabulary that survives a provider's copyright filter, and how to check the length, the aspect ratio and the mix of what came back. Use when a piece has more than one scene and the video model writes its own audio (Seedance 2, Veo 3, MiniMax H3, Kling 2.6 and later), or when a cut's sound drops out at the shot changes.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fvideo-audio-continuity\u002FSKILL.md",[180],{"categoryNo":120,"categoryName":121,"taxonomyType":86,"taxonomyTypeDesc":87},[182,142,141,183,184,185,186,187,27,188,189,190,191,25,192,193,140,194,195,196,197],"video audio continuity","continuity","write","why","back","veo 3","cut","later","across","filter","generated","how to","mix","design","boundary","how"]