[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6BY27J519CQXZCE090S60PC":3,"skill-same:01M6BY27J519CQXZCE090S60PC":70},{"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":36,"linkOnly":37,"markdownContent":38,"markdownBody":39,"frontMatter":40,"refName":42,"commitSha":43,"skillDirectory":44,"forksCount":45,"repositoryHtmlUrl":46,"repositoryHomepageUrl":47,"ownerHtmlUrl":48,"ownerBlogUrl":12,"ownerDescription":49,"updatedAt":50,"install":51,"resources":63},"01M6BY27J519CQXZCE090S60PC","nodetool-js-scripting",null,"Write JavaScript that runs in NodeTool's QuickJS sandbox: Code node bodies, saved JS script documents, sandbox package imports, and calling nodes from code.","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\u002Fnodetool-js-scripting\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35],"scripting","js","nodetool js scripting","nodetool","imports","sandbox","true","saved","script","calling","Nodes","write","node","package","code","documents","javascript","2026-11-10 10:12:30",false,"---\nname: nodetool-js-scripting\ndescription: \"Write JavaScript that runs in NodeTool's QuickJS sandbox: Code node bodies, saved JS script documents, sandbox package imports, and calling nodes from code.\"\nfeatured: true\n---\n\n# Write NodeTool sandbox JavaScript\n\nThree places take the same guest JavaScript, with the same globals, the same\nimports and the same limits.\n\n| Where | What it is | Reach it with |\n|---|---|---|\n| **Code node** | A body inside a workflow graph | `validate_code`, `run_code`, `test_code` |\n| **JS script document** | A named, versioned script with declared ports, secrets, a timeout and saved tests | `list_js_scripts`, `get_js_script`, `save_js_script`, `edit_js_script`, `validate_js_script`, `run_js_script`, `test_js_script` |\n| **Native flow** | Guest code calling nodes as async functions, no graph | `import \"@nodetool-ai\u002Fsandbox-nodetool\u002Fflow\"` |\n\nPick a Code node when the logic belongs to one graph. Pick a script document\nwhen the same logic is called from several places, needs its own tests, or needs\nversion history. Pick native flow when the control flow is easier to write than\nto wire.\n\n## The body contract\n\n```js\n\u002F\u002F inputs: { rows: [...], threshold: 10 }\nconst kept = inputs.rows.filter((r) => r.score > inputs.threshold);\nprogress(50, `kept ${kept.length}`);\nfor (const row of kept) await emit(\"row\", row);   \u002F\u002F streams as it goes\nawait output(\"count\", kept.length);               \u002F\u002F final, posted at the end\n```\n\n- **Inputs arrive on the `inputs` object**, never as globals of their own name.\n  Values are deep-copied through JSON before entering the guest.\n- **Outputs leave through `await output(name, value)` and `await emit(name,\n  value)`, and nothing else.** `name` must be a declared output handle. A second\n  `output` on the same handle throws. `return` is ordinary control flow and its\n  value is ignored. For a script document this is enforced: a body that declares\n  outputs and returns them instead of emitting them is an **error**, because a\n  script has no legacy return contract. A Code node still accepts the old\n  return\u002Fyield contract for one more release, with a deprecation warning.\n- **Media inputs arrive as refs.** Read one with `media.bytes` \u002F `media.text` \u002F\n  `media.info`, and build one with `media.toDocument` \u002F `toImage` \u002F `toAudio` \u002F\n  `toVideo`. Pass the whole input object, not its `uri`, so the ref's own type\n  travels with it. Every `media.*` call is async.\n- **`stream` is the input side of `emit`.** A body that mentions `stream` runs\n  **once** over the whole stream and pulls its own items. A body that never\n  mentions it runs once per incoming item. Nothing is configured: deleting the\n  last `stream` call flips the node back.\n\n```js\nfor await (const item of stream(name))            \u002F\u002F one handle, in order, until EOS\nfor await (const [handle, item] of stream.any())  \u002F\u002F every handle, arrival order\nconst item = await stream.first(name)             \u002F\u002F next value, undefined at EOS\nstream.open(name)                                 \u002F\u002F could more still arrive?\n```\n\n- **`state`** survives across streaming invocations and resets at each run.\n- **`progress(percent, message)`** drives the node's progress bar.\n\n## Capabilities are globals, libraries are imports\n\nGlobals need no declaration: `console`, `fetch`, `workspace`, `getSecret`,\n`nodetool.secrets.*`, `sleep`, `crypto`, `format`, `image`, `audio`, `video`,\n`media`, `canvas`, `assetToSandbox` \u002F `sandboxToAsset`, `progress`, `emit` \u002F\n`output`, and the pure helpers `toBase64` \u002F `fromBase64` \u002F `toHex` \u002F `fromHex` \u002F\n`parallelMap`. `media.*` and `workspace.*` need a `ProcessingContext`, so they\nthrow in a bare sandbox call with no run behind it.\n\nEverything else is a **sandbox package** the body imports:\n\n```js\nimport yaml from \"@nodetool-ai\u002Fsandbox-yaml\";\n```\n\nDiscover them with `list_sandbox_packages` and read one's API with\n`get_sandbox_package_docs`. Do not guess a specifier: a pack this host does not\ncarry fails validation with \"Install `\u003Cpack>`\". `nodetool packs\ncompile` builds one for a dependency that is not shipped.\n\n`image.*`, `audio.*` and `video.*` transforms return a `sandbox:\u002F\u002Fmedia\u002F\u003Cid>`\nhandle, not bytes. Chain the calls, then `\u003Ctype>.toAsset(handle)` before the run\nends. A handle does not survive into a later run.\n\n## Running an agent from code\n\nUse `run_agent` when the work needs more than one model call: look something\nup, act, read the result, correct.\n\n```js\nimport { run_agent } from \"@nodetool-ai\u002Fsandbox-nodetool\u002Fagents\";\n\nconst answer = await run_agent({\n  prompt: \"Read storyboard \" + inputs.storyboardId + \" and list its shot titles.\",\n  model: inputs.model,                \u002F\u002F {provider, id} from a model selector\n  tools: [\"get_storyboard\"],          \u002F\u002F capability wire names, nothing else\n  output_schema: { type: \"object\", properties: { titles: { type: \"array\", items: { type: \"string\" } } } }\n});\nawait output(\"titles\", answer.result.titles);\n```\n\nWithout `output_schema`, `answer.text` is the agent's final message. The agent\nreports its text and tool calls while it runs. A mini app shows them in an\nAgent Activity widget bound to `op:\u003Cid>\u002Fexec#transcript`. `generate_text` stays\nthe right call for one prompt with no tools.\n\n## Calling nodes from code\n\n```js\nimport \"@nodetool-ai\u002Fsandbox-nodetool\u002Fflow\";   \u002F\u002F mounts the bridge, required\nimport { concat } from \"@nodetool-ai\u002Fsandbox-flow\u002Fnodetool.text\";\n\nconst r = await concat({ a: inputs.left, b: inputs.right });\nawait output(\"joined\", r.output);\n```\n\n`await` is the edge, a variable is the wire, `Promise.all` is the fan-out. Both\nimports are required: the facade does not mount without the capability module.\nStreaming-output nodes carry `.stream(inputs)`, an async iterable where an early\n`break` closes the stream and runs node cleanup. Every call passes the per-call\npermission gate and bills through the invoking run, bounded by a recursion depth\ncap of 4 and 16 concurrently open streams per run. Streaming *inputs* accept\narrays only in v1.\n\nA program that must open in the editor, be validated, or run on the server still\nbuilds a graph. Use `nodetool-workflow-builder`\nfor that.\n\n## Building a timeline from code\n\n```js\nimport { video } from \"@nodetool-ai\u002Fsandbox-timeline\";\n\nconst v = video({ width: 1920, height: 1080, fps: 30 });\nconst title = v.scene(\"title\", 3, (s) => {\n  const t = s.text(\"Hello\", { size: 120, weight: 600, color: \"#ffffff\" });\n  t.enter({ from: { opacity: 0 }, at: 0, dur: 0.5 });\n});\nv.series([title]);\nconst saved = await v.save(nodetool.timelines, { name: \"Hello\" });\nawait output(\"timeline_id\", saved.timeline_id);\n```\n\nThe pack authors a whole motion-graphics cut in seconds, local to each scene,\nand saves it with one `set_timeline_document`, instead of one `edit_timeline`\nop per clip. Pass the body's `nodetool.timelines` to `v.save`: a module cannot\nsee the belt. `get_sandbox_package_docs` on the specifier returns the full\nauthoring API.\n\nSave the build script itself as a JS script document and it links to the\ntimeline it makes automatically: `v.save(nodetool.timelines, {name})` with no\n`timeline_id` the first time creates the timeline and links this script to\nit; every later run of the same script updates that timeline in place\ninstead of making another. Revise with `edit_js_script`'s string-replacement\nops (below) rather than resending the whole body.\n\n## The loop\n\n1. **Write or read the body.** `get_js_script` for a saved script.\n2. **`validate_code` \u002F `validate_js_script` after every edit.** It catches\n   syntax, imports against the installed catalog, undefined names, undeclared\n   `inputs.*` reads, outputs no `emit`\u002F`output` call reaches, duplicate or\n   non-identifier port names, and tests naming ports the script does not\n   declare. It is far cheaper than running.\n3. **`run_code` \u002F `run_js_script`** with real inputs. `output` values come back\n   as `outputs`, `emit` values as `streamed`, an ordered list of `{name, value}`.\n   Stage a streamed input with `input_streams`, keyed by handle.\n4. **`test_code` \u002F `test_js_script`** as the regression check. A case supplies\n   `inputs` (or `input_streams`) and optionally `expect` — final values per\n   handle, compared structurally, with unnamed outputs ignored, and\n   `expected_streamed`, the full ordered emit list. A case with neither passes\n   when the body runs without error.\n5. **Save.** `save_js_script` validates first and is CAS on update. For a\n   small revision, `edit_js_script(js_script_id, ops)` is cheaper: each op is\n   `{old, new}`, and `old` must match exactly once in the current code — like\n   a code editor's find\u002Freplace, refused by name on zero or on an ambiguous\n   match. It validates the result the same way `save_js_script` does.\n\nScript documents get the same version family as the other documents:\n`list_js_script_versions`, `get_js_script_version`, `create_js_script_version`,\n`restore_js_script_version`, `delete_js_script_version`.\n\n## Secrets and limits\n\nRead a credential with `nodetool.secrets.get(name)` — never inline one, and\nnever write one into a body or a test case. A script's declared secrets are\nintersected with whatever allowance the invoking context carries. There is no\n`set_secret`: `request_secret` asks the user's own client for one and the value\nnever enters the guest, the transcript, or the model's context. A headless run\ncarries no secret prompt, so the call is refused by name rather than quietly\nwriting something nobody approved.\n\nScript composition is bounded like sub-agents: depth cap 4 with a script id\nchain, so a cycle fails the call naming it. Execution time defaults to 30s and\nevery limit is overridable per invocation and clamped to a ceiling.\n\n## Verify from a shell\n\nThese are the `nodetool` CLI. From a NodeTool checkout the same commands run\nas `npm run dev:nodetool -- \u003Ccommand>`.\n\n```bash\nnodetool jsscript validate \u003Cid|file.json> --json\nnodetool jsscript run \u003Cid|file.json> --inputs '{\"numbers\":[1,2,3]}'\nnodetool jsscript run \u003Cid|file.json> --input-streams '{\"numbers\":[1,2,3]}'\nnodetool jsscript test \u003Cid|file.json> --json\nnodetool jsscript versions list|show|create|restore|delete \u003Cid>\n```\n\nA path that exists on disk wins over an id, and a file target needs no database.\n`jsscript test` exits non-zero on any failure and is the keyless selfcheck the\nharness gate runs.\n\n## Reference\n\nIn a NodeTool checkout, these repository sources go further:\n\n- `docs\u002Fjavascript-sandbox.md` — the guest surface, marshaling, limits,\n  concurrency and the security model.\n- `docs\u002Fjs-script-document-design.md` — document shape, storage, invocation\n  from agents, Code nodes and mini apps.\n- `docs\u002Fharnesses.md` § nodetool jsscript\n- `packages\u002Fsandbox-packs\u002FREADME.md`\n","# Write NodeTool sandbox JavaScript\n\nThree places take the same guest JavaScript, with the same globals, the same\nimports and the same limits.\n\n| Where | What it is | Reach it with |\n|---|---|---|\n| **Code node** | A body inside a workflow graph | `validate_code`, `run_code`, `test_code` |\n| **JS script document** | A named, versioned script with declared ports, secrets, a timeout and saved tests | `list_js_scripts`, `get_js_script`, `save_js_script`, `edit_js_script`, `validate_js_script`, `run_js_script`, `test_js_script` |\n| **Native flow** | Guest code calling nodes as async functions, no graph | `import \"@nodetool-ai\u002Fsandbox-nodetool\u002Fflow\"` |\n\nPick a Code node when the logic belongs to one graph. Pick a script document\nwhen the same logic is called from several places, needs its own tests, or needs\nversion history. Pick native flow when the control flow is easier to write than\nto wire.\n\n## The body contract\n\n```js\n\u002F\u002F inputs: { rows: [...], threshold: 10 }\nconst kept = inputs.rows.filter((r) => r.score > inputs.threshold);\nprogress(50, `kept ${kept.length}`);\nfor (const row of kept) await emit(\"row\", row);   \u002F\u002F streams as it goes\nawait output(\"count\", kept.length);               \u002F\u002F final, posted at the end\n```\n\n- **Inputs arrive on the `inputs` object**, never as globals of their own name.\n  Values are deep-copied through JSON before entering the guest.\n- **Outputs leave through `await output(name, value)` and `await emit(name,\n  value)`, and nothing else.** `name` must be a declared output handle. A second\n  `output` on the same handle throws. `return` is ordinary control flow and its\n  value is ignored. For a script document this is enforced: a body that declares\n  outputs and returns them instead of emitting them is an **error**, because a\n  script has no legacy return contract. A Code node still accepts the old\n  return\u002Fyield contract for one more release, with a deprecation warning.\n- **Media inputs arrive as refs.** Read one with `media.bytes` \u002F `media.text` \u002F\n  `media.info`, and build one with `media.toDocument` \u002F `toImage` \u002F `toAudio` \u002F\n  `toVideo`. Pass the whole input object, not its `uri`, so the ref's own type\n  travels with it. Every `media.*` call is async.\n- **`stream` is the input side of `emit`.** A body that mentions `stream` runs\n  **once** over the whole stream and pulls its own items. A body that never\n  mentions it runs once per incoming item. Nothing is configured: deleting the\n  last `stream` call flips the node back.\n\n```js\nfor await (const item of stream(name))            \u002F\u002F one handle, in order, until EOS\nfor await (const [handle, item] of stream.any())  \u002F\u002F every handle, arrival order\nconst item = await stream.first(name)             \u002F\u002F next value, undefined at EOS\nstream.open(name)                                 \u002F\u002F could more still arrive?\n```\n\n- **`state`** survives across streaming invocations and resets at each run.\n- **`progress(percent, message)`** drives the node's progress bar.\n\n## Capabilities are globals, libraries are imports\n\nGlobals need no declaration: `console`, `fetch`, `workspace`, `getSecret`,\n`nodetool.secrets.*`, `sleep`, `crypto`, `format`, `image`, `audio`, `video`,\n`media`, `canvas`, `assetToSandbox` \u002F `sandboxToAsset`, `progress`, `emit` \u002F\n`output`, and the pure helpers `toBase64` \u002F `fromBase64` \u002F `toHex` \u002F `fromHex` \u002F\n`parallelMap`. `media.*` and `workspace.*` need a `ProcessingContext`, so they\nthrow in a bare sandbox call with no run behind it.\n\nEverything else is a **sandbox package** the body imports:\n\n```js\nimport yaml from \"@nodetool-ai\u002Fsandbox-yaml\";\n```\n\nDiscover them with `list_sandbox_packages` and read one's API with\n`get_sandbox_package_docs`. Do not guess a specifier: a pack this host does not\ncarry fails validation with \"Install `\u003Cpack>`\". `nodetool packs\ncompile` builds one for a dependency that is not shipped.\n\n`image.*`, `audio.*` and `video.*` transforms return a `sandbox:\u002F\u002Fmedia\u002F\u003Cid>`\nhandle, not bytes. Chain the calls, then `\u003Ctype>.toAsset(handle)` before the run\nends. A handle does not survive into a later run.\n\n## Running an agent from code\n\nUse `run_agent` when the work needs more than one model call: look something\nup, act, read the result, correct.\n\n```js\nimport { run_agent } from \"@nodetool-ai\u002Fsandbox-nodetool\u002Fagents\";\n\nconst answer = await run_agent({\n  prompt: \"Read storyboard \" + inputs.storyboardId + \" and list its shot titles.\",\n  model: inputs.model,                \u002F\u002F {provider, id} from a model selector\n  tools: [\"get_storyboard\"],          \u002F\u002F capability wire names, nothing else\n  output_schema: { type: \"object\", properties: { titles: { type: \"array\", items: { type: \"string\" } } } }\n});\nawait output(\"titles\", answer.result.titles);\n```\n\nWithout `output_schema`, `answer.text` is the agent's final message. The agent\nreports its text and tool calls while it runs. A mini app shows them in an\nAgent Activity widget bound to `op:\u003Cid>\u002Fexec#transcript`. `generate_text` stays\nthe right call for one prompt with no tools.\n\n## Calling nodes from code\n\n```js\nimport \"@nodetool-ai\u002Fsandbox-nodetool\u002Fflow\";   \u002F\u002F mounts the bridge, required\nimport { concat } from \"@nodetool-ai\u002Fsandbox-flow\u002Fnodetool.text\";\n\nconst r = await concat({ a: inputs.left, b: inputs.right });\nawait output(\"joined\", r.output);\n```\n\n`await` is the edge, a variable is the wire, `Promise.all` is the fan-out. Both\nimports are required: the facade does not mount without the capability module.\nStreaming-output nodes carry `.stream(inputs)`, an async iterable where an early\n`break` closes the stream and runs node cleanup. Every call passes the per-call\npermission gate and bills through the invoking run, bounded by a recursion depth\ncap of 4 and 16 concurrently open streams per run. Streaming *inputs* accept\narrays only in v1.\n\nA program that must open in the editor, be validated, or run on the server still\nbuilds a graph. Use `nodetool-workflow-builder`\nfor that.\n\n## Building a timeline from code\n\n```js\nimport { video } from \"@nodetool-ai\u002Fsandbox-timeline\";\n\nconst v = video({ width: 1920, height: 1080, fps: 30 });\nconst title = v.scene(\"title\", 3, (s) => {\n  const t = s.text(\"Hello\", { size: 120, weight: 600, color: \"#ffffff\" });\n  t.enter({ from: { opacity: 0 }, at: 0, dur: 0.5 });\n});\nv.series([title]);\nconst saved = await v.save(nodetool.timelines, { name: \"Hello\" });\nawait output(\"timeline_id\", saved.timeline_id);\n```\n\nThe pack authors a whole motion-graphics cut in seconds, local to each scene,\nand saves it with one `set_timeline_document`, instead of one `edit_timeline`\nop per clip. Pass the body's `nodetool.timelines` to `v.save`: a module cannot\nsee the belt. `get_sandbox_package_docs` on the specifier returns the full\nauthoring API.\n\nSave the build script itself as a JS script document and it links to the\ntimeline it makes automatically: `v.save(nodetool.timelines, {name})` with no\n`timeline_id` the first time creates the timeline and links this script to\nit; every later run of the same script updates that timeline in place\ninstead of making another. Revise with `edit_js_script`'s string-replacement\nops (below) rather than resending the whole body.\n\n## The loop\n\n1. **Write or read the body.** `get_js_script` for a saved script.\n2. **`validate_code` \u002F `validate_js_script` after every edit.** It catches\n   syntax, imports against the installed catalog, undefined names, undeclared\n   `inputs.*` reads, outputs no `emit`\u002F`output` call reaches, duplicate or\n   non-identifier port names, and tests naming ports the script does not\n   declare. It is far cheaper than running.\n3. **`run_code` \u002F `run_js_script`** with real inputs. `output` values come back\n   as `outputs`, `emit` values as `streamed`, an ordered list of `{name, value}`.\n   Stage a streamed input with `input_streams`, keyed by handle.\n4. **`test_code` \u002F `test_js_script`** as the regression check. A case supplies\n   `inputs` (or `input_streams`) and optionally `expect` — final values per\n   handle, compared structurally, with unnamed outputs ignored, and\n   `expected_streamed`, the full ordered emit list. A case with neither passes\n   when the body runs without error.\n5. **Save.** `save_js_script` validates first and is CAS on update. For a\n   small revision, `edit_js_script(js_script_id, ops)` is cheaper: each op is\n   `{old, new}`, and `old` must match exactly once in the current code — like\n   a code editor's find\u002Freplace, refused by name on zero or on an ambiguous\n   match. It validates the result the same way `save_js_script` does.\n\nScript documents get the same version family as the other documents:\n`list_js_script_versions`, `get_js_script_version`, `create_js_script_version`,\n`restore_js_script_version`, `delete_js_script_version`.\n\n## Secrets and limits\n\nRead a credential with `nodetool.secrets.get(name)` — never inline one, and\nnever write one into a body or a test case. A script's declared secrets are\nintersected with whatever allowance the invoking context carries. There is no\n`set_secret`: `request_secret` asks the user's own client for one and the value\nnever enters the guest, the transcript, or the model's context. A headless run\ncarries no secret prompt, so the call is refused by name rather than quietly\nwriting something nobody approved.\n\nScript composition is bounded like sub-agents: depth cap 4 with a script id\nchain, so a cycle fails the call naming it. Execution time defaults to 30s and\nevery limit is overridable per invocation and clamped to a ceiling.\n\n## Verify from a shell\n\nThese are the `nodetool` CLI. From a NodeTool checkout the same commands run\nas `npm run dev:nodetool -- \u003Ccommand>`.\n\n```bash\nnodetool jsscript validate \u003Cid|file.json> --json\nnodetool jsscript run \u003Cid|file.json> --inputs '{\"numbers\":[1,2,3]}'\nnodetool jsscript run \u003Cid|file.json> --input-streams '{\"numbers\":[1,2,3]}'\nnodetool jsscript test \u003Cid|file.json> --json\nnodetool jsscript versions list|show|create|restore|delete \u003Cid>\n```\n\nA path that exists on disk wins over an id, and a file target needs no database.\n`jsscript test` exits non-zero on any failure and is the keyless selfcheck the\nharness gate runs.\n\n## Reference\n\nIn a NodeTool checkout, these repository sources go further:\n\n- `docs\u002Fjavascript-sandbox.md` — the guest surface, marshaling, limits,\n  concurrency and the security model.\n- `docs\u002Fjs-script-document-design.md` — document shape, storage, invocation\n  from agents, Code nodes and mini apps.\n- `docs\u002Fharnesses.md` § nodetool jsscript\n- `packages\u002Fsandbox-packs\u002FREADME.md`\n",{"name":5,"featured":41,"description":7},true,"main","c7226baa82daca42a19a5855c116544f97aa8bac","packages\u002Fsystem-skills\u002Fnodetool-js-scripting",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":52,"prompt":62,"zipAvailable":41},[53,56,59],{"tool":54,"command":55},"npx","npx skills add nodetool-ai\u002Fnodetool --skill nodetool-js-scripting",{"tool":57,"command":58},"bunx","bunx skills add nodetool-ai\u002Fnodetool --skill nodetool-js-scripting",{"tool":60,"command":61},"pnpm","pnpm dlx skills add nodetool-ai\u002Fnodetool --skill nodetool-js-scripting","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fnodetool-ai\u002Fnodetool\u002Fblob\u002Fmain\u002Fpackages\u002Fsystem-skills\u002Fnodetool-js-scripting\u002FSKILL.md",[64],{"path":65,"resourceType":66,"resourceTypeDesc":67,"sizeBytes":68,"sourceHtmlUrl":69},"prompt.md","REFERENCE","参考文档",9186,"https:\u002F\u002Fgithub.com\u002Fshenxingy\u002FClade\u002Fblob\u002Fmain\u002Fconfigs\u002Fskills\u002Fads-competitor\u002Fprompt.md",[71,80,116,146,161,177],{"skillNo":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":75,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":79},"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":81,"name":22,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":95,"publishedAt":36},"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",[85,90],{"categoryNo":86,"categoryName":87,"taxonomyType":88,"taxonomyTypeDesc":89},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",{"categoryNo":91,"categoryName":92,"taxonomyType":93,"taxonomyTypeDesc":94},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[22,96,97,98,99,100,101,102,103,104,105,106,31,107,108,109,110,111,112,113,114,115],"MCP","Media","workflows","when","generation","media generation","with","visual","use","collections","for","work","debug","graph","validate","assets","run","server","mcp server","workflow",{"skillNo":117,"name":118,"nameZh":6,"description":119,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":120,"categories":121,"keywords":125,"publishedAt":36},"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",[122],{"categoryNo":123,"categoryName":124,"taxonomyType":88,"taxonomyTypeDesc":89},"01JBSEEDCAT000000000000016","图像与多媒体",[126,127,128,129,130,131,132,31,133,134,135,136,137,138,139,140,141,142,143,144,145],"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":147,"name":148,"nameZh":6,"description":149,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":150,"categories":151,"keywords":154,"publishedAt":36},"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",[152,153],{"categoryNo":123,"categoryName":124,"taxonomyType":88,"taxonomyTypeDesc":89},{"categoryNo":86,"categoryName":87,"taxonomyType":88,"taxonomyTypeDesc":89},[115,155,145,156,157,158,159,104,160,22,106],"video workflow","reusable","storyboard","build","batches","single",{"skillNo":162,"name":163,"nameZh":6,"description":164,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":165,"categories":166,"keywords":168,"publishedAt":36},"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",[167],{"categoryNo":123,"categoryName":124,"taxonomyType":88,"taxonomyTypeDesc":89},[169,170,145,102,171,172,22,173,174,175,176],"clone","video clone","cast","new","analyze","product","structure","reference",{"skillNo":178,"name":179,"nameZh":6,"description":180,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":181,"categories":182,"keywords":184,"publishedAt":36},"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",[183],{"categoryNo":123,"categoryName":124,"taxonomyType":88,"taxonomyTypeDesc":89},[185,145,144,186,30,187,188,189,190,191,192,193,194,99,195,196,143,197,198,199,200],"video audio continuity","continuity","why","back","veo 3","at","cut","later","across","filter","generated","how to","mix","design","boundary","how"]