[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6BY2E5PAYNEJRXQJVMFMFF7":3,"skill-same:01M6BY2E5PAYNEJRXQJVMFMFF7":73},{"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":23,"publishedAt":45,"linkOnly":46,"markdownContent":47,"markdownBody":48,"frontMatter":49,"refName":51,"commitSha":52,"skillDirectory":53,"forksCount":54,"repositoryHtmlUrl":55,"repositoryHomepageUrl":56,"ownerHtmlUrl":57,"ownerBlogUrl":12,"ownerDescription":58,"updatedAt":59,"install":60,"resources":72},"01M6BY2E5PAYNEJRXQJVMFMFF7","nodetool-rag-indexing",null,"Build NodeTool document ingestion, vector indexing, retrieval, and RAG pipelines. Use when documents must be chunked into a vector collection and searched by meaning, or when an answer should come from the user's own documents. To find a stored image, video or file by name or tag, use asset_search instead.","nodetool-ai\u002Fnodetool","nodetool-ai","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F164873569?v=4","01M6BXPFDW36K9K1MR8ZH63HG2","nodetool.ai","01M6BXPFJ893XM219GQN60B0WK",633,"AGPL-3.0","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fnodetool-rag-indexing\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44],"rag","nodetool rag indexing","nodetool","indexing","use","by name","when","true","meaning","answer","Document Ingestion","asset","vector indexing","ingestion","image","Video","collection","file","vector","search","documents","2026-11-10 10:12:30",false,"---\nname: nodetool-rag-indexing\ndescription: \"Build NodeTool document ingestion, vector indexing, retrieval, and RAG pipelines. Use when documents must be chunked into a vector collection and searched by meaning, or when an answer should come from the user's own documents. To find a stored image, video or file by name or tag, use asset_search instead.\"\nfeatured: true\n---\n\n# Index documents and answer from them\n\nA **collection** is a named vector store. Indexing splits documents into\nchunks, embeds each chunk with the collection's embedding model and stores it.\nRetrieval embeds the question with the same model and returns the closest\nchunks, which a language model answers from.\n\n```\nINDEX:  documents → text → chunks → vector.IndexTextChunk → collection\nQUERY:  question → vector.HybridSearch \u002F QueryText → context → Agent → answer\n```\n\n## Search what is already indexed\n\nTwo tools are on the agent's belt in every chat:\n\n- `list_collections` returns each collection the user can access, with its\n  metadata (including `embedding_model`).\n- `query_collection {collection, query, n_results?}` returns the closest chunks\n  as `{id, document, score}`. `n_results` defaults to 5.\n\nAnswer from the returned chunks and say which chunk ids the answer used. An\nempty result means the collection holds nothing close, not that the topic is\nunknown.\n\nThe `vector_index`, `vector_batch_index`, `vector_text_search`,\n`vector_hybrid_search`, `vector_recursive_split_and_index` and\n`vector_markdown_split_and_index` tools act only on a collection a host binds\nto the run. A chat binds none, so they return an error there. Index through the\nnodes below instead.\n\n## Index through nodes\n\nAll vector nodes are in the `vector` namespace.\n\n| Node | Takes | Does |\n|---|---|---|\n| `vector.Collection` | `name`, `embedding_model` | Gets or creates the collection and outputs a `{type: \"collection\", name}` ref |\n| `vector.IndexTextChunk` | `collection`, `document_id`, `text`, `metadata` | Upserts one chunk. The same `document_id` overwrites |\n| `vector.IndexString` | `collection`, `text`, `document_id`, `metadata` | Upserts one string |\n| `vector.IndexImage`, `vector.IndexEmbedding` | | Index an image or a precomputed embedding |\n| `vector.QueryText` | `collection`, `text`, `n_results` | Vector search. Outputs `ids`, `documents`, `metadatas`, `distances` |\n| `vector.HybridSearch` | `collection`, `text`, `n_results`, `k_constant`, `min_keyword_length` | Vector plus keyword search, fused by reciprocal rank. Adds `scores` |\n| `vector.QueryImage` | `collection`, `image`, `n_results` | Image search |\n| `vector.GetDocuments`, `vector.Peek`, `vector.Count` | `collection` | Inspect what is stored |\n| `vector.RemoveOverlap` | `documents` | Removes text repeated between overlapping chunks |\n\nThe embedding model is set once, on `vector.Collection`'s `embedding_model`,\nand stored in the collection's metadata. Index and query nodes take only the\ncollection, so indexing and search always use the same model. The shipped\nexample uses the Ollama model `nomic-embed-text`. Changing the model means a\nnew collection.\n\nFor a few chunks, `run_node` is enough: `run_node {node_type:\n\"vector.Collection\", inputs: {name, embedding_model}}` once, then\n`vector.IndexTextChunk` with `collection: {type: \"collection\", name}` per\nchunk. For a folder of documents, build a workflow and run it.\n\n### Index workflow\n\n1. **List files.** A Code node with `await workspace.list(dir)`, or\n   `nodetool.document.ListDocuments {folder, pattern, recursive}`.\n2. **Load text.** `nodetool.document.LoadDocumentFile {path}` for text and\n   Markdown. For PDFs use `lib.pdf.ExtractText` or `lib.pdf.ExtractMarkdown`.\n3. **Chunk.** There is no chunking node. Use a Code node, starting from the\n   \"Chunk Text\" snippet, and record the list with\n   `await output(\"chunks\", chunks)`. Read a document input with\n   `await media.text(inputs.document)`.\n4. **Fan out.** `nodetool.control.ForEach` emits each chunk on `output` and its\n   position on `index`. Build a unique `document_id` from the file name and the\n   index, for example with `nodetool.text.Template`.\n5. **Index.** `vector.IndexTextChunk` with the collection ref, the chunk as\n   `text`, the id as `document_id`, and `{source: \u003Cfile name>}` as `metadata`.\n\nValidate with `validate_workflow`, then run it. Indexing with a hosted\nembedding model is billed per token, so confirm the size of the corpus with the\nuser before indexing a large folder.\n\n| Content | Chunk size | Overlap |\n|---|---|---|\n| Technical docs | 200–500 tokens | about 50 tokens |\n| Prose | 300–600 tokens | about 75 tokens |\n| Code | 100–300 tokens | about 25 tokens |\n\n### Query workflow\n\n`StringInput` (question) → `vector.HybridSearch {collection, n_results: 5}` →\na Code node joining `documents` into one context string → `nodetool.text.Prompt`\nor `Template` with the context and the question → `nodetool.agents.Agent` with a\nsystem prompt that limits it to the context → `Output`.\n\nThe shipped example **Chat With Your Documents** builds a collection, indexes\nthree documents and answers from them. Load it with `get_example_workflow\n{package_name: \"nodetool-base\", example_name: \"Chat With Your Documents\"}` and\nadapt it rather than starting from nothing.\n\n## Storage backends\n\nThe workflow nodes and tools are the same on every backend. The server picks\none at start from `NODETOOL_VECTOR_PROVIDER`:\n\n| Value | Store | Configuration |\n|---|---|---|\n| `sqlite-vec` (default) | Local SQLite file | `VECTORSTORE_DB_PATH` to move it |\n| `pinecone` | Pinecone | `PINECONE_API_KEY`, optional `PINECONE_ENVIRONMENT`, `PINECONE_CONTROLLER_HOST` |\n| `supabase` | Postgres with pgvector | `SUPABASE_URL` and `SUPABASE_KEY` or `SUPABASE_SERVICE_ROLE_KEY`, plus `packages\u002Fvectorstore\u002Fsql\u002Fsupabase-migration.sql` applied to the project |\n\nAny other value stops the server with an error. These are server environment\nvariables, set by whoever runs the server, not by the agent.\n\n## From a shell or over HTTP\n\nFor a person with a shell, the `nodetool collections` CLI manages collections\non the local server:\n\n```bash\nnodetool collections create handbook --embedding-model nomic-embed-text --embedding-provider ollama\nnodetool collections index handbook .\u002Fdocs\u002F*.md\nnodetool collections query handbook \"refund policy\" -n 5\nnodetool collections list\nnodetool collections delete handbook --yes\n```\n\n`POST \u002Fapi\u002Fcollections\u002F\u003Cname>\u002Findex` takes one text file as a\n`multipart\u002Fform-data` upload in the `file` field, chunks it and indexes it into\nan existing collection that has an embedding model. Plain text, Markdown, HTML,\nCSV and JSON are accepted. PDFs, Office files and images need an indexing\nworkflow.\n\n```bash\ncurl -X POST http:\u002F\u002Flocalhost:7777\u002Fapi\u002Fcollections\u002Fhandbook\u002Findex \\\n  -H \"Authorization: Bearer $TOKEN\" \\\n  -F \"file=@.\u002Fdocs\u002Frefunds.md\"\n```\n\n## Pitfalls\n\n- **A collection with no embedding model** cannot be searched. Create it with\n  one.\n- **Chunks too large** dilute the context. **Chunks too small** split\n  sentences. Keep 10–20% overlap.\n- **Reused `document_id`s** overwrite each other. Include the file name in the\n  id.\n- **Querying before indexing** returns nothing.\n- **Node names**: `vector.IndexTextChunk`, `vector.QueryText`,\n  `vector.HybridSearch`. There is no `vector.chroma.*` or `vector.faiss.*`, and\n  no `nodetool.text.Chunk`.\n","# Index documents and answer from them\n\nA **collection** is a named vector store. Indexing splits documents into\nchunks, embeds each chunk with the collection's embedding model and stores it.\nRetrieval embeds the question with the same model and returns the closest\nchunks, which a language model answers from.\n\n```\nINDEX:  documents → text → chunks → vector.IndexTextChunk → collection\nQUERY:  question → vector.HybridSearch \u002F QueryText → context → Agent → answer\n```\n\n## Search what is already indexed\n\nTwo tools are on the agent's belt in every chat:\n\n- `list_collections` returns each collection the user can access, with its\n  metadata (including `embedding_model`).\n- `query_collection {collection, query, n_results?}` returns the closest chunks\n  as `{id, document, score}`. `n_results` defaults to 5.\n\nAnswer from the returned chunks and say which chunk ids the answer used. An\nempty result means the collection holds nothing close, not that the topic is\nunknown.\n\nThe `vector_index`, `vector_batch_index`, `vector_text_search`,\n`vector_hybrid_search`, `vector_recursive_split_and_index` and\n`vector_markdown_split_and_index` tools act only on a collection a host binds\nto the run. A chat binds none, so they return an error there. Index through the\nnodes below instead.\n\n## Index through nodes\n\nAll vector nodes are in the `vector` namespace.\n\n| Node | Takes | Does |\n|---|---|---|\n| `vector.Collection` | `name`, `embedding_model` | Gets or creates the collection and outputs a `{type: \"collection\", name}` ref |\n| `vector.IndexTextChunk` | `collection`, `document_id`, `text`, `metadata` | Upserts one chunk. The same `document_id` overwrites |\n| `vector.IndexString` | `collection`, `text`, `document_id`, `metadata` | Upserts one string |\n| `vector.IndexImage`, `vector.IndexEmbedding` | | Index an image or a precomputed embedding |\n| `vector.QueryText` | `collection`, `text`, `n_results` | Vector search. Outputs `ids`, `documents`, `metadatas`, `distances` |\n| `vector.HybridSearch` | `collection`, `text`, `n_results`, `k_constant`, `min_keyword_length` | Vector plus keyword search, fused by reciprocal rank. Adds `scores` |\n| `vector.QueryImage` | `collection`, `image`, `n_results` | Image search |\n| `vector.GetDocuments`, `vector.Peek`, `vector.Count` | `collection` | Inspect what is stored |\n| `vector.RemoveOverlap` | `documents` | Removes text repeated between overlapping chunks |\n\nThe embedding model is set once, on `vector.Collection`'s `embedding_model`,\nand stored in the collection's metadata. Index and query nodes take only the\ncollection, so indexing and search always use the same model. The shipped\nexample uses the Ollama model `nomic-embed-text`. Changing the model means a\nnew collection.\n\nFor a few chunks, `run_node` is enough: `run_node {node_type:\n\"vector.Collection\", inputs: {name, embedding_model}}` once, then\n`vector.IndexTextChunk` with `collection: {type: \"collection\", name}` per\nchunk. For a folder of documents, build a workflow and run it.\n\n### Index workflow\n\n1. **List files.** A Code node with `await workspace.list(dir)`, or\n   `nodetool.document.ListDocuments {folder, pattern, recursive}`.\n2. **Load text.** `nodetool.document.LoadDocumentFile {path}` for text and\n   Markdown. For PDFs use `lib.pdf.ExtractText` or `lib.pdf.ExtractMarkdown`.\n3. **Chunk.** There is no chunking node. Use a Code node, starting from the\n   \"Chunk Text\" snippet, and record the list with\n   `await output(\"chunks\", chunks)`. Read a document input with\n   `await media.text(inputs.document)`.\n4. **Fan out.** `nodetool.control.ForEach` emits each chunk on `output` and its\n   position on `index`. Build a unique `document_id` from the file name and the\n   index, for example with `nodetool.text.Template`.\n5. **Index.** `vector.IndexTextChunk` with the collection ref, the chunk as\n   `text`, the id as `document_id`, and `{source: \u003Cfile name>}` as `metadata`.\n\nValidate with `validate_workflow`, then run it. Indexing with a hosted\nembedding model is billed per token, so confirm the size of the corpus with the\nuser before indexing a large folder.\n\n| Content | Chunk size | Overlap |\n|---|---|---|\n| Technical docs | 200–500 tokens | about 50 tokens |\n| Prose | 300–600 tokens | about 75 tokens |\n| Code | 100–300 tokens | about 25 tokens |\n\n### Query workflow\n\n`StringInput` (question) → `vector.HybridSearch {collection, n_results: 5}` →\na Code node joining `documents` into one context string → `nodetool.text.Prompt`\nor `Template` with the context and the question → `nodetool.agents.Agent` with a\nsystem prompt that limits it to the context → `Output`.\n\nThe shipped example **Chat With Your Documents** builds a collection, indexes\nthree documents and answers from them. Load it with `get_example_workflow\n{package_name: \"nodetool-base\", example_name: \"Chat With Your Documents\"}` and\nadapt it rather than starting from nothing.\n\n## Storage backends\n\nThe workflow nodes and tools are the same on every backend. The server picks\none at start from `NODETOOL_VECTOR_PROVIDER`:\n\n| Value | Store | Configuration |\n|---|---|---|\n| `sqlite-vec` (default) | Local SQLite file | `VECTORSTORE_DB_PATH` to move it |\n| `pinecone` | Pinecone | `PINECONE_API_KEY`, optional `PINECONE_ENVIRONMENT`, `PINECONE_CONTROLLER_HOST` |\n| `supabase` | Postgres with pgvector | `SUPABASE_URL` and `SUPABASE_KEY` or `SUPABASE_SERVICE_ROLE_KEY`, plus `packages\u002Fvectorstore\u002Fsql\u002Fsupabase-migration.sql` applied to the project |\n\nAny other value stops the server with an error. These are server environment\nvariables, set by whoever runs the server, not by the agent.\n\n## From a shell or over HTTP\n\nFor a person with a shell, the `nodetool collections` CLI manages collections\non the local server:\n\n```bash\nnodetool collections create handbook --embedding-model nomic-embed-text --embedding-provider ollama\nnodetool collections index handbook .\u002Fdocs\u002F*.md\nnodetool collections query handbook \"refund policy\" -n 5\nnodetool collections list\nnodetool collections delete handbook --yes\n```\n\n`POST \u002Fapi\u002Fcollections\u002F\u003Cname>\u002Findex` takes one text file as a\n`multipart\u002Fform-data` upload in the `file` field, chunks it and indexes it into\nan existing collection that has an embedding model. Plain text, Markdown, HTML,\nCSV and JSON are accepted. PDFs, Office files and images need an indexing\nworkflow.\n\n```bash\ncurl -X POST http:\u002F\u002Flocalhost:7777\u002Fapi\u002Fcollections\u002Fhandbook\u002Findex \\\n  -H \"Authorization: Bearer $TOKEN\" \\\n  -F \"file=@.\u002Fdocs\u002Frefunds.md\"\n```\n\n## Pitfalls\n\n- **A collection with no embedding model** cannot be searched. Create it with\n  one.\n- **Chunks too large** dilute the context. **Chunks too small** split\n  sentences. Keep 10–20% overlap.\n- **Reused `document_id`s** overwrite each other. Include the file name in the\n  id.\n- **Querying before indexing** returns nothing.\n- **Node names**: `vector.IndexTextChunk`, `vector.QueryText`,\n  `vector.HybridSearch`. There is no `vector.chroma.*` or `vector.faiss.*`, and\n  no `nodetool.text.Chunk`.\n",{"name":5,"featured":50,"description":7},true,"main","c7226baa82daca42a19a5855c116544f97aa8bac","packages\u002Fsystem-skills\u002Fnodetool-rag-indexing",90,"https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool","https:\u002F\u002Fnodetool.ai\u002F","https:\u002F\u002Fgithub.com\u002Fnodetool-ai","No-code AI development","2026-11-14 03:12:08",{"source":8,"skillName":5,"commands":61,"prompt":71,"zipAvailable":50},[62,65,68],{"tool":63,"command":64},"npx","npx skills add nodetool-ai\u002Fnodetool --skill nodetool-rag-indexing",{"tool":66,"command":67},"bunx","bunx skills add nodetool-ai\u002Fnodetool --skill nodetool-rag-indexing",{"tool":69,"command":70},"pnpm","pnpm dlx skills add nodetool-ai\u002Fnodetool --skill nodetool-rag-indexing","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fnodetool-rag-indexing\u002FSKILL.md",[],[74,107,133,148,164,189],{"skillNo":75,"name":26,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":87,"publishedAt":45},"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",[79,82],{"categoryNo":80,"categoryName":81,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",{"categoryNo":83,"categoryName":84,"taxonomyType":85,"taxonomyTypeDesc":86},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[26,88,89,90,91,92,93,94,95,96,97,98,99,100,28,101,102,103,104,105,106,30],"MCP","mcp server","workflow","workflows","validate","debug","run","graph","server","Media","assets","work","node","visual","generation","collections","media generation","with","for",{"skillNo":108,"name":109,"nameZh":6,"description":110,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":111,"categories":112,"keywords":114,"publishedAt":45},"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",[113],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[115,116,38,117,118,39,119,120,121,122,123,100,124,125,126,127,128,129,130,131,132],"wan 2 6 prompting","prompting","generate","Budget","Audio","text","animate","model","flash","only","route","time","line","input","text-to-video","image-to-video","mode","expansion",{"skillNo":134,"name":135,"nameZh":6,"description":136,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":137,"categories":138,"keywords":141,"publishedAt":45},"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",[139,140],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},{"categoryNo":80,"categoryName":81,"taxonomyType":21,"taxonomyTypeDesc":22},[90,39,142,26,28,143,144,145,146,106,147],"video workflow","reusable","storyboard","build","batches","single",{"skillNo":149,"name":150,"nameZh":6,"description":151,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":152,"categories":153,"keywords":155,"publishedAt":45},"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",[154],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[39,156,157,158,159,26,160,161,162,105,163],"video clone","clone","product","analyze","reference","cast","structure","new",{"skillNo":165,"name":166,"nameZh":6,"description":167,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":168,"categories":169,"keywords":171,"publishedAt":45},"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",[170],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[39,119,172,173,174,175,176,177,122,178,179,180,181,182,183,184,185,186,187,188,30],"video audio continuity","continuity","design","boundary","how","why","write","generated","later","back","across","cut","how to","filter","mix","at","veo 3",{"skillNo":190,"name":191,"nameZh":6,"description":192,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":193,"categories":194,"keywords":196,"publishedAt":45},"01M6BY3ZANBSQMZRH4MG55R8Z7","veo-3-prompting","Prompt Google's Veo 3 video line — the five-element shot\u002Fsetting\u002Fsubject\u002Faction\u002Fdialogue structure, the cinematography vocabulary it reads directly, duration and aspect-ratio tradeoffs, the audio and auto-fix switches per provider, and the fixes for visual drift, temporal inconsistency and audio mismatch. Use whenever the model id contains veo3 or veo-3 (fal-ai\u002Fveo3.1 and its fast, lite, image-to-video, first-last-frame and reference-to-video routes, fal-ai\u002Fveo3\u002Fimage-to-video, google\u002Fveo3.1 on atlascloud, google\u002Fveo-3 and google\u002Fveo-3.1 on replicate, kie's veo-3-1, and Gemini's own veo-3.1-generate-preview) on generate_video, animate_image, generate_video_from_references or a TextToVideo node.","https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fveo-3-prompting\u002FSKILL.md",[195],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},[197,198,116,188,38,199,200,117,201,202,39,119,121,122,203,204,205,206,207,208,209,162,210,211,212],"veo 3 prompting","veo","google","gemini","fast","first","fal","kie","replicate","fix","auto","references","drift","preview","element","auto-fix"]