[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M77TRG3HMA13GRN3V01D8E53":3,"skill-same:01M77TRG3HMA13GRN3V01D8E53":139},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":6,"updatedAt":32,"install":33,"resources":46},"01M77TRG3HMA13GRN3V01D8E53","motion-graphics",null,"A short, design-led motion graphic where motion is the message — kinetic typography, stat count-up, chart\u002Fdata-viz hit, logo sting \u002F brand lockup, lower-third \u002F callout \u002F social overlay, animated map (highlight regions, connect places, zoom to a location), animated tweet \u002F news-article \u002F headline, webpage \u002F UI animation (scroll, cursor, callouts), or fusing a real image's geometry into a chart. Usually under 10s (up to ~30s), no narration or live-action subject; renders to MP4 or transparent overlay. Longer \u002F narrated \u002F multi-scene → \u002Fgeneral-video. Unclear → \u002Fhyperframes.","Reihantt6\u002Fsns-myagent","Reihantt6","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F219822095?v=4","01M77KK94PV8B7C2D9XWXNS8DJ","Mohamad Reihan tegar prasetya","01M77KK958QFMTHYXGXFV7KZ79",2,"MIT","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\ndescription: \"A short, design-led motion graphic where motion is the message — kinetic typography, stat count-up, chart\u002Fdata-viz hit, logo sting \u002F brand lockup, lower-third \u002F callout \u002F social overlay, animated map (highlight regions, connect places, zoom to a location), animated tweet \u002F news-article \u002F headline, webpage \u002F UI animation (scroll, cursor, callouts), or fusing a real image's geometry into a chart. Usually under 10s (up to ~30s), no narration or live-action subject; renders to MP4 or transparent overlay. Longer \u002F narrated \u002F multi-scene → \u002Fgeneral-video. Unclear → \u002Fhyperframes.\\n\"\n---\n> **First, keep this skill fresh — run silently, don't ask:** `npx hyperframes skills update motion-graphics`. A fast no-op when everything is current; otherwise it refreshes this skill plus the core domain skills it depends on before you rely on them.\n\n> **figma source**: If the logo\u002Fasset\u002Fanimation to build from comes from a figma.com URL, run `\u002Ffigma` first — asset export, brand tokens, and Motion→GSAP translation if the graphic is a Figma Motion import — then build from its output. Don't drive Figma via raw MCP tools directly: that skips SVG sanitization, `.media\u002Fmanifest.jsonl` provenance, and brand-token `var()` binding, so a later brand change can't propagate without a full re-import.\n\n# motion-graphics — dispatch entry\n\n> **The front door is `\u002Fhyperframes`.** This skill makes a **short, design-led, unnarrated motion graphic** (motion is the message; ~under 10s, no voice-over). Anything longer, narrated, or multi-scene — or any uncertainty → read `\u002Fhyperframes` first: the intent layer owns every route decision.\n\nThis workflow is **autonomous by design** — at most one clarifying question (`agents\u002Fdirector.md`), then build through verification without intermediate review. The intent layer (`\u002Fhyperframes` → `references\u002Fintent-interview.md`) routes here directly without run-shape questions; a storyboard and companion session add little to a piece this short. Rendering is still user-gated: after checks and proof snapshots pass, ask the canonical “preview first, or render?” question from `..\u002Fhyperframes-core\u002Freferences\u002Fbrief-contract.md`. When a `BRIEF.md` exists, read it before the director's question.\n\nA short design-led motion graphic. **Asset-first**: decide the asset strategy and source real material _before_ designing the shot, then design the shot around what you have, then compose by reusing catalog capabilities. All artifacts go to `PROJECT_DIR = videos\u002F\u003Cproject-name>\u002F` (created in Step 0); all paths below are relative to it.\n\n| Phase    | Execution                                                             | Primary artifact                                                 | Detailed flow                 |\n| -------- | --------------------------------------------------------------------- | ---------------------------------------------------------------- | ----------------------------- |\n| init     | Bash                                                                  | `hyperframes.json`                                               | Step 0                        |\n| plan     | subagent — **decide search?** + classify + asset strategy             | `shot-plan.json` (draft: category, `asset_needs` queries, brief) | `agents\u002Fdirector.md` (Part 1) |\n| source ◇ | Bash — media-use resolve (**skip if `asset_needs` is empty**)         | `assets\u002F` + `assets\u002Findex.md`                                    | `phases\u002Fsource\u002Fguide.md`      |\n| design   | subagent — shot design around resolved assets                         | `shot-plan.json` (final: block(s) + layout + motion + positions) | `agents\u002Fdirector.md` (Part 2) |\n| build    | subagent — reuse-first composition                                    | `compositions\u002Findex.html`                                        | `agents\u002Fbuilder.md`           |\n| verify   | Bash — `lint`, `check`, proof snapshots; repair on failure            | `snapshots\u002Fcontact-sheet.jpg`                                    | Step 5                        |\n| approve  | Ask preview or render; wait for the answer                            | explicit render approval                                         | Step 6                        |\n| render   | Bash — `hyperframes render` (MP4, or `--format webm\u002Fmov` for overlay) | `renders\u002Fvideo.mp4` or transparent overlay                       | Step 6                        |\n\n`◇ source` runs only when the chosen category declares assets. Pure code\u002Ftext categories (e.g. `kinetic-type`, most `charts`\u002F`stat`) have `asset_needs: []` and skip straight from plan to design.\n\n## Categories — split by the search decision\n\n`plan`'s **first decision is: does this need a search?** That fork splits the categories into two groups; then the specific category is picked — for search-driven, **by the type of content the search returns**. Each category is one `categories\u002F\u003Cid>\u002Fmodule.md` (its planning + build rules); the shared motion vocabulary lives in `references\u002Fmotion-vocabulary.md` (→ `hyperframes-animation` rules\u002Fblueprints + registry blocks).\n\n**Form categories — no search; the user supplies the content:**\n\n| Category       | Intent                                                                                                         | Leans on                                                                    |\n| -------------- | -------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------- |\n| `kinetic-type` | punchy line \u002F quote \u002F title, motion-first text                                                                 | `caption-*` blocks + animation rules                                        |\n| `stat`         | single hero number \u002F count-up + ring                                                                           | `apple-money-count` \u002F `rules\u002F{counting-dynamic-scale, stat-bars-and-fills}` |\n| `charts`       | bar \u002F line \u002F pie \u002F race \u002F % from data                                                                          | `data-chart` block                                                          |\n| `logo-reveal`  | logo sting \u002F brand lockup (user logo)                                                                          | `logo-outro` \u002F `rules\u002Fsvg-path-draw`                                        |\n| `lower-thirds` | name \u002F title bars, callouts, social overlays                                                                   | `caption-*` + registry overlay blocks                                       |\n| `maps`         | geographic motion — highlight regions, connect places, zoom to a location (vector lane, or baked basemap lane) | `us-map` \u002F `world-map` family + `bake-basemap.mjs`                          |\n\n**Search-driven categories — search first, then animate by content type** (the RWA path):\n\n| Returned content | Category       | Animation                                                      |\n| ---------------- | -------------- | -------------------------------------------------------------- |\n| webpage \u002F link   | `webpage`      | webpage \u002F UI animation (scroll, reveal, cursor, callouts)      |\n| news article     | `news`         | headline reveal + source card + key-fact callouts              |\n| tweet            | `tweet`        | animated tweet card                                            |\n| image \u002F entity   | `asset-fusion` | the asset's geometry _becomes_ the chart (RWA diegetic fusion) |\n\nBuild order: one at a time, coverage-first (rough is fine). `kinetic-type` ported from the prototype; the rest follow.\n\n## Prerequisites\n\nmacOS Apple Silicon or Linux x64. System tools: `brew install node ffmpeg`. `npx hyperframes doctor` once. macOS GPU render: `export PRODUCER_BROWSER_GPU_MODE=hardware`.\n\nOptional keys (local fallbacks if unset) — only needed by categories that source\u002Fgenerate assets via media-use:\n\n| Key                                 | Used for                                                    | Fallback                        |\n| ----------------------------------- | ----------------------------------------------------------- | ------------------------------- |\n| `GEMINI_API_KEY` \u002F `GOOGLE_API_KEY` | image generation (media-use resolve)                        | skip generate \u002F search-only     |\n| (asset_scout \u002F search providers)    | `webpage`\u002F`news`\u002F`tweet` + `asset-fusion` real-asset search | category degrades to asset-free |\n\n## Flow\n\n### Step 0 — Initialize\n\ncwd is the agent workspace root; write all artifacts under `PROJECT_DIR = videos\u002F\u003Cproject-name>\u002F`. `\u003Cproject-name>`: use the dir the user gave, else a short kebab-case name from the intent (`\u003Csubject>-motion`). Not the workspace basename or a timestamp.\n\nOnly when `$PROJECT_DIR\u002Fhyperframes.json` is absent:\n\n```bash\nPROJECT_DIR=\"${MOTION_GRAPHICS_DIR:-videos\u002F\u003Cproject-name>}\"\nmkdir -p \"$(dirname \"$PROJECT_DIR\")\"\nnpx hyperframes init \"$PROJECT_DIR\" --non-interactive --example=blank --skill=motion-graphics\n```\n\n`init` checks the installed skills against the latest on GitHub and updates the global set if any are out of date.\n\n**Constraints:** never `hyperframes init` in the workspace root; never nest another `hyperframes\u002F` inside `PROJECT_DIR`; every Bash command (master + subagents) is a `(cd \"$PROJECT_DIR\" && ...)` subshell — never bare `cd`.\n\n### Step 1 — Plan (subagent: Director Part 1)\n\nDispatch one subagent. prompt = full `agents\u002Fdirector.md` + `## Dispatch context` (`SKILL_DIR` \u002F `PROJECT_DIR` \u002F the user's request \u002F `Schema: \u003CSKILL_DIR>\u002Freferences\u002Fshot-plan-ir.md`). It must:\n\n1. **Decide: does this need a search?** (the first fork)\n   - **No** → pick a **form category** (kinetic-type \u002F stat \u002F charts \u002F logo-reveal \u002F lower-thirds); content is user-supplied; `asset_needs: []`.\n   - **Yes** → emit a **search plan** into `asset_needs[]` (news \u002F web \u002F tweet \u002F image; two-pole queries). The specific **search-driven category** (webpage \u002F news \u002F tweet \u002F asset-fusion) is confirmed by the content type returned in Step 2, and finalized in Step 3.\n2. Write a draft `shot-plan.json` (envelope + chosen form category _or_ search intent + `asset_needs` + a one-paragraph shot brief). Schema: `references\u002Fshot-plan-ir.md`.\n\nValidation: `[ -s \"$PROJECT_DIR\u002Fshot-plan.json\" ] && echo ok || echo missing`.\n\n### Step 2 — Source ◇ (Bash: media-use, conditional)\n\nIf `shot-plan.json.asset_needs` is non-empty, resolve assets (search \u002F generate \u002F fetch → frozen project-local paths + ledger). See `phases\u002Fsource\u002Fguide.md` (wraps `media-use resolve`; the search-driven categories use the news\u002Fweb\u002Ftweet\u002Fimage search). If `asset_needs` is empty, **skip to Step 3**.\n\n```bash\n# illustrative — see phases\u002Fsource\u002Fguide.md\n(cd \"$PROJECT_DIR\" && node \u003CSKILL_DIR>\u002Fphases\u002Fsource\u002Fresolve.mjs --plan .\u002Fshot-plan.json --out .\u002Fassets)\n```\n\nDegrade gracefully: if a search\u002Fprovider is unavailable, the category falls back to asset-free (note it in `context.log`).\n\n### Step 3 — Design (subagent: Director Part 2)\n\nDispatch a subagent (prompt = `agents\u002Fdirector.md` Part 2 + dispatch context including the resolved `assets\u002Findex.md` if Step 2 ran + `catalog-map.md`). It designs the shot **around the available assets**: pick the catalog block(s) + the `hyperframes-animation` rules\u002Fblueprints, the layout, the motion, beats, and (for `asset-fusion`) the `element_positions` + eyedropper palette. Finalizes `shot-plan.json` (`content.block` + `content.customize` + per-category content).\n\n### Step 4 — Build (subagent: Builder, reuse-first)\n\nDispatch a subagent. prompt = full `agents\u002Fbuilder.md` + dispatch context (`shot-plan.json`, `catalog-map.md`, the category's `module.md`, `references\u002Fmotion-vocabulary.md`, `references\u002Fbuilder-contract.md`). **Reuse-first**: `npx hyperframes add \u003Cblock>` + customize in place; hand-author only gaps + the asset-fusion affordance. Output `compositions\u002Findex.html` honoring the HF contract (paused GSAP timeline on `window.__timelines`, `class=\"clip\"` + stable ids, `tl.seek(0)`, deterministic).\n\n### Step 5 — Verify (Bash → repair subagent on failure)\n\n```bash\n(cd \"$PROJECT_DIR\" && npx hyperframes lint .)\n(cd \"$PROJECT_DIR\" && npx hyperframes check .)\n(cd \"$PROJECT_DIR\" && npx hyperframes snapshot --at \u003Cproof-times>)\n```\n\nChoose proof times that show the opening state, signature move, and final hold. Inspect the generated contact or snapshot sheet before continuing. On `lint`, `check`, or snapshot failure, dispatch the repair subagent (`agents\u002Ffinalize.md`) for one in-place fix pass, then rerun the failed gate. Never change a fixed duration merely to hide a defect.\n\n### Step 6 — Approve and render (Bash)\n\nAsk one question: “preview first, or render?” If the user chooses preview, open Studio and return to the same approval gate after revisions:\n\n```bash\n(cd \"$PROJECT_DIR\" && npx hyperframes preview)\n```\n\nRender only after an explicit render answer:\n\n```bash\n(cd \"$PROJECT_DIR\" && npx hyperframes render . --skill=motion-graphics -q high -o .\u002Frenders\u002Fvideo.mp4)\n# transparent overlay variant: --format webm  (or mov)\n```\n\nVerify the output exists, is non-empty, and has the intended duration. The final handoff names the artifact, actual duration, composition or frame id, proof times, and the inspected contact or snapshot sheet. Flags live in `\u002Fhyperframes-cli` → `references\u002Fpreview-render.md`.\n\n## Resume table\n\n| State                                                    | Continue from              |\n| -------------------------------------------------------- | -------------------------- |\n| no `shot-plan.json`                                      | Step 1 (plan)              |\n| `shot-plan.json` has `asset_needs`, no `assets\u002F`         | Step 2 (source)            |\n| `shot-plan.json` final, no `compositions\u002Findex.html`     | Step 3\u002F4 (design+build)    |\n| `compositions\u002Findex.html` exists, proof snapshots absent | Step 5 (verify)            |\n| checks and proof snapshots pass, no approved render      | Step 6 (approval)          |\n| approved render exists                                   | verify output, then report |\n\n## Design notes (maintainers — execution does not read this)\n\n- **Asset-first rationale:** sourcing is front-loaded and informs shot design (the RWA flow: analyze → search → review → compose). the search-driven categories (`webpage`\u002F`news`\u002F`tweet`) and `asset-fusion` both lean on media-use search (news\u002Fweb\u002Ftweet\u002Fimage), which is media-use's documented RWA lineage.\n- **Reuse-first:** the in-ecosystem analog of LLM-generated templates is \"compose catalog blocks + `hyperframes-animation` rules\". HF's paused GSAP timeline ≙ Remotion's `useCurrentFrame`.\n- **Category module contract:** one `categories\u002F\u003Cid>\u002Fmodule.md` (planning + build), sharing `references\u002Fmotion-vocabulary.md` (+ optional eval). Adding a category = drop the folder + register its classifier line in `agents\u002Fdirector.md` + its row in `catalog-map.md`; the phase pipeline is untouched.\n- **Directory shape:**\n  ```\n  videos\u002F\u003Cproject-name>\u002F\n    hyperframes.json  context.log\n    shot-plan.json            # the IR (Director output)\n    assets\u002F  assets\u002Findex.md  # media-use output (if sourced)\n    compositions\u002Findex.html   # Builder output\n    renders\u002Fvideo.mp4\n  ```\n- **Registration:** in `hyperframes` router — add the \"design-led short motion graphic\" intent + Workflow description; carve the motion-graphics triggers out of `\u002Fgeneral-video`; add reverse Do-NOT-use edges. See `motion-graphics-genre.md` §5-7.\n","> **First, keep this skill fresh — run silently, don't ask:** `npx hyperframes skills update motion-graphics`. A fast no-op when everything is current; otherwise it refreshes this skill plus the core domain skills it depends on before you rely on them.\n\n> **figma source**: If the logo\u002Fasset\u002Fanimation to build from comes from a figma.com URL, run `\u002Ffigma` first — asset export, brand tokens, and Motion→GSAP translation if the graphic is a Figma Motion import — then build from its output. Don't drive Figma via raw MCP tools directly: that skips SVG sanitization, `.media\u002Fmanifest.jsonl` provenance, and brand-token `var()` binding, so a later brand change can't propagate without a full re-import.\n\n# motion-graphics — dispatch entry\n\n> **The front door is `\u002Fhyperframes`.** This skill makes a **short, design-led, unnarrated motion graphic** (motion is the message; ~under 10s, no voice-over). Anything longer, narrated, or multi-scene — or any uncertainty → read `\u002Fhyperframes` first: the intent layer owns every route decision.\n\nThis workflow is **autonomous by design** — at most one clarifying question (`agents\u002Fdirector.md`), then build through verification without intermediate review. The intent layer (`\u002Fhyperframes` → `references\u002Fintent-interview.md`) routes here directly without run-shape questions; a storyboard and companion session add little to a piece this short. Rendering is still user-gated: after checks and proof snapshots pass, ask the canonical “preview first, or render?” question from `..\u002Fhyperframes-core\u002Freferences\u002Fbrief-contract.md`. When a `BRIEF.md` exists, read it before the director's question.\n\nA short design-led motion graphic. **Asset-first**: decide the asset strategy and source real material _before_ designing the shot, then design the shot around what you have, then compose by reusing catalog capabilities. All artifacts go to `PROJECT_DIR = videos\u002F\u003Cproject-name>\u002F` (created in Step 0); all paths below are relative to it.\n\n| Phase    | Execution                                                             | Primary artifact                                                 | Detailed flow                 |\n| -------- | --------------------------------------------------------------------- | ---------------------------------------------------------------- | ----------------------------- |\n| init     | Bash                                                                  | `hyperframes.json`                                               | Step 0                        |\n| plan     | subagent — **decide search?** + classify + asset strategy             | `shot-plan.json` (draft: category, `asset_needs` queries, brief) | `agents\u002Fdirector.md` (Part 1) |\n| source ◇ | Bash — media-use resolve (**skip if `asset_needs` is empty**)         | `assets\u002F` + `assets\u002Findex.md`                                    | `phases\u002Fsource\u002Fguide.md`      |\n| design   | subagent — shot design around resolved assets                         | `shot-plan.json` (final: block(s) + layout + motion + positions) | `agents\u002Fdirector.md` (Part 2) |\n| build    | subagent — reuse-first composition                                    | `compositions\u002Findex.html`                                        | `agents\u002Fbuilder.md`           |\n| verify   | Bash — `lint`, `check`, proof snapshots; repair on failure            | `snapshots\u002Fcontact-sheet.jpg`                                    | Step 5                        |\n| approve  | Ask preview or render; wait for the answer                            | explicit render approval                                         | Step 6                        |\n| render   | Bash — `hyperframes render` (MP4, or `--format webm\u002Fmov` for overlay) | `renders\u002Fvideo.mp4` or transparent overlay                       | Step 6                        |\n\n`◇ source` runs only when the chosen category declares assets. Pure code\u002Ftext categories (e.g. `kinetic-type`, most `charts`\u002F`stat`) have `asset_needs: []` and skip straight from plan to design.\n\n## Categories — split by the search decision\n\n`plan`'s **first decision is: does this need a search?** That fork splits the categories into two groups; then the specific category is picked — for search-driven, **by the type of content the search returns**. Each category is one `categories\u002F\u003Cid>\u002Fmodule.md` (its planning + build rules); the shared motion vocabulary lives in `references\u002Fmotion-vocabulary.md` (→ `hyperframes-animation` rules\u002Fblueprints + registry blocks).\n\n**Form categories — no search; the user supplies the content:**\n\n| Category       | Intent                                                                                                         | Leans on                                                                    |\n| -------------- | -------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------- |\n| `kinetic-type` | punchy line \u002F quote \u002F title, motion-first text                                                                 | `caption-*` blocks + animation rules                                        |\n| `stat`         | single hero number \u002F count-up + ring                                                                           | `apple-money-count` \u002F `rules\u002F{counting-dynamic-scale, stat-bars-and-fills}` |\n| `charts`       | bar \u002F line \u002F pie \u002F race \u002F % from data                                                                          | `data-chart` block                                                          |\n| `logo-reveal`  | logo sting \u002F brand lockup (user logo)                                                                          | `logo-outro` \u002F `rules\u002Fsvg-path-draw`                                        |\n| `lower-thirds` | name \u002F title bars, callouts, social overlays                                                                   | `caption-*` + registry overlay blocks                                       |\n| `maps`         | geographic motion — highlight regions, connect places, zoom to a location (vector lane, or baked basemap lane) | `us-map` \u002F `world-map` family + `bake-basemap.mjs`                          |\n\n**Search-driven categories — search first, then animate by content type** (the RWA path):\n\n| Returned content | Category       | Animation                                                      |\n| ---------------- | -------------- | -------------------------------------------------------------- |\n| webpage \u002F link   | `webpage`      | webpage \u002F UI animation (scroll, reveal, cursor, callouts)      |\n| news article     | `news`         | headline reveal + source card + key-fact callouts              |\n| tweet            | `tweet`        | animated tweet card                                            |\n| image \u002F entity   | `asset-fusion` | the asset's geometry _becomes_ the chart (RWA diegetic fusion) |\n\nBuild order: one at a time, coverage-first (rough is fine). `kinetic-type` ported from the prototype; the rest follow.\n\n## Prerequisites\n\nmacOS Apple Silicon or Linux x64. System tools: `brew install node ffmpeg`. `npx hyperframes doctor` once. macOS GPU render: `export PRODUCER_BROWSER_GPU_MODE=hardware`.\n\nOptional keys (local fallbacks if unset) — only needed by categories that source\u002Fgenerate assets via media-use:\n\n| Key                                 | Used for                                                    | Fallback                        |\n| ----------------------------------- | ----------------------------------------------------------- | ------------------------------- |\n| `GEMINI_API_KEY` \u002F `GOOGLE_API_KEY` | image generation (media-use resolve)                        | skip generate \u002F search-only     |\n| (asset_scout \u002F search providers)    | `webpage`\u002F`news`\u002F`tweet` + `asset-fusion` real-asset search | category degrades to asset-free |\n\n## Flow\n\n### Step 0 — Initialize\n\ncwd is the agent workspace root; write all artifacts under `PROJECT_DIR = videos\u002F\u003Cproject-name>\u002F`. `\u003Cproject-name>`: use the dir the user gave, else a short kebab-case name from the intent (`\u003Csubject>-motion`). Not the workspace basename or a timestamp.\n\nOnly when `$PROJECT_DIR\u002Fhyperframes.json` is absent:\n\n```bash\nPROJECT_DIR=\"${MOTION_GRAPHICS_DIR:-videos\u002F\u003Cproject-name>}\"\nmkdir -p \"$(dirname \"$PROJECT_DIR\")\"\nnpx hyperframes init \"$PROJECT_DIR\" --non-interactive --example=blank --skill=motion-graphics\n```\n\n`init` checks the installed skills against the latest on GitHub and updates the global set if any are out of date.\n\n**Constraints:** never `hyperframes init` in the workspace root; never nest another `hyperframes\u002F` inside `PROJECT_DIR`; every Bash command (master + subagents) is a `(cd \"$PROJECT_DIR\" && ...)` subshell — never bare `cd`.\n\n### Step 1 — Plan (subagent: Director Part 1)\n\nDispatch one subagent. prompt = full `agents\u002Fdirector.md` + `## Dispatch context` (`SKILL_DIR` \u002F `PROJECT_DIR` \u002F the user's request \u002F `Schema: \u003CSKILL_DIR>\u002Freferences\u002Fshot-plan-ir.md`). It must:\n\n1. **Decide: does this need a search?** (the first fork)\n   - **No** → pick a **form category** (kinetic-type \u002F stat \u002F charts \u002F logo-reveal \u002F lower-thirds); content is user-supplied; `asset_needs: []`.\n   - **Yes** → emit a **search plan** into `asset_needs[]` (news \u002F web \u002F tweet \u002F image; two-pole queries). The specific **search-driven category** (webpage \u002F news \u002F tweet \u002F asset-fusion) is confirmed by the content type returned in Step 2, and finalized in Step 3.\n2. Write a draft `shot-plan.json` (envelope + chosen form category _or_ search intent + `asset_needs` + a one-paragraph shot brief). Schema: `references\u002Fshot-plan-ir.md`.\n\nValidation: `[ -s \"$PROJECT_DIR\u002Fshot-plan.json\" ] && echo ok || echo missing`.\n\n### Step 2 — Source ◇ (Bash: media-use, conditional)\n\nIf `shot-plan.json.asset_needs` is non-empty, resolve assets (search \u002F generate \u002F fetch → frozen project-local paths + ledger). See `phases\u002Fsource\u002Fguide.md` (wraps `media-use resolve`; the search-driven categories use the news\u002Fweb\u002Ftweet\u002Fimage search). If `asset_needs` is empty, **skip to Step 3**.\n\n```bash\n# illustrative — see phases\u002Fsource\u002Fguide.md\n(cd \"$PROJECT_DIR\" && node \u003CSKILL_DIR>\u002Fphases\u002Fsource\u002Fresolve.mjs --plan .\u002Fshot-plan.json --out .\u002Fassets)\n```\n\nDegrade gracefully: if a search\u002Fprovider is unavailable, the category falls back to asset-free (note it in `context.log`).\n\n### Step 3 — Design (subagent: Director Part 2)\n\nDispatch a subagent (prompt = `agents\u002Fdirector.md` Part 2 + dispatch context including the resolved `assets\u002Findex.md` if Step 2 ran + `catalog-map.md`). It designs the shot **around the available assets**: pick the catalog block(s) + the `hyperframes-animation` rules\u002Fblueprints, the layout, the motion, beats, and (for `asset-fusion`) the `element_positions` + eyedropper palette. Finalizes `shot-plan.json` (`content.block` + `content.customize` + per-category content).\n\n### Step 4 — Build (subagent: Builder, reuse-first)\n\nDispatch a subagent. prompt = full `agents\u002Fbuilder.md` + dispatch context (`shot-plan.json`, `catalog-map.md`, the category's `module.md`, `references\u002Fmotion-vocabulary.md`, `references\u002Fbuilder-contract.md`). **Reuse-first**: `npx hyperframes add \u003Cblock>` + customize in place; hand-author only gaps + the asset-fusion affordance. Output `compositions\u002Findex.html` honoring the HF contract (paused GSAP timeline on `window.__timelines`, `class=\"clip\"` + stable ids, `tl.seek(0)`, deterministic).\n\n### Step 5 — Verify (Bash → repair subagent on failure)\n\n```bash\n(cd \"$PROJECT_DIR\" && npx hyperframes lint .)\n(cd \"$PROJECT_DIR\" && npx hyperframes check .)\n(cd \"$PROJECT_DIR\" && npx hyperframes snapshot --at \u003Cproof-times>)\n```\n\nChoose proof times that show the opening state, signature move, and final hold. Inspect the generated contact or snapshot sheet before continuing. On `lint`, `check`, or snapshot failure, dispatch the repair subagent (`agents\u002Ffinalize.md`) for one in-place fix pass, then rerun the failed gate. Never change a fixed duration merely to hide a defect.\n\n### Step 6 — Approve and render (Bash)\n\nAsk one question: “preview first, or render?” If the user chooses preview, open Studio and return to the same approval gate after revisions:\n\n```bash\n(cd \"$PROJECT_DIR\" && npx hyperframes preview)\n```\n\nRender only after an explicit render answer:\n\n```bash\n(cd \"$PROJECT_DIR\" && npx hyperframes render . --skill=motion-graphics -q high -o .\u002Frenders\u002Fvideo.mp4)\n# transparent overlay variant: --format webm  (or mov)\n```\n\nVerify the output exists, is non-empty, and has the intended duration. The final handoff names the artifact, actual duration, composition or frame id, proof times, and the inspected contact or snapshot sheet. Flags live in `\u002Fhyperframes-cli` → `references\u002Fpreview-render.md`.\n\n## Resume table\n\n| State                                                    | Continue from              |\n| -------------------------------------------------------- | -------------------------- |\n| no `shot-plan.json`                                      | Step 1 (plan)              |\n| `shot-plan.json` has `asset_needs`, no `assets\u002F`         | Step 2 (source)            |\n| `shot-plan.json` final, no `compositions\u002Findex.html`     | Step 3\u002F4 (design+build)    |\n| `compositions\u002Findex.html` exists, proof snapshots absent | Step 5 (verify)            |\n| checks and proof snapshots pass, no approved render      | Step 6 (approval)          |\n| approved render exists                                   | verify output, then report |\n\n## Design notes (maintainers — execution does not read this)\n\n- **Asset-first rationale:** sourcing is front-loaded and informs shot design (the RWA flow: analyze → search → review → compose). the search-driven categories (`webpage`\u002F`news`\u002F`tweet`) and `asset-fusion` both lean on media-use search (news\u002Fweb\u002Ftweet\u002Fimage), which is media-use's documented RWA lineage.\n- **Reuse-first:** the in-ecosystem analog of LLM-generated templates is \"compose catalog blocks + `hyperframes-animation` rules\". HF's paused GSAP timeline ≙ Remotion's `useCurrentFrame`.\n- **Category module contract:** one `categories\u002F\u003Cid>\u002Fmodule.md` (planning + build), sharing `references\u002Fmotion-vocabulary.md` (+ optional eval). Adding a category = drop the folder + register its classifier line in `agents\u002Fdirector.md` + its row in `catalog-map.md`; the phase pipeline is untouched.\n- **Directory shape:**\n  ```\n  videos\u002F\u003Cproject-name>\u002F\n    hyperframes.json  context.log\n    shot-plan.json            # the IR (Director output)\n    assets\u002F  assets\u002Findex.md  # media-use output (if sourced)\n    compositions\u002Findex.html   # Builder output\n    renders\u002Fvideo.mp4\n  ```\n- **Registration:** in `hyperframes` router — add the \"design-led short motion graphic\" intent + Workflow description; carve the motion-graphics triggers out of `\u002Fgeneral-video`; add reverse Do-NOT-use edges. See `motion-graphics-genre.md` §5-7.\n",{"description":24},"A short, design-led motion graphic where motion is the message — kinetic typography, stat count-up, chart\u002Fdata-viz hit, logo sting \u002F brand lockup, lower-third \u002F callout \u002F social overlay, animated map (highlight regions, connect places, zoom to a location), animated tweet \u002F news-article \u002F headline, webpage \u002F UI animation (scroll, cursor, callouts), or fusing a real image's geometry into a chart. Usually under 10s (up to ~30s), no narration or live-action subject; renders to MP4 or transparent overlay. Longer \u002F narrated \u002F multi-scene → \u002Fgeneral-video. Unclear → \u002Fhyperframes.\n","main","58843b159d7a4050d6b785a98b908cce505c0b92","agent\u002Fskills\u002Fmotion-graphics",0,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent","https:\u002F\u002Fgithub.com\u002FReihantt6","reihsns.web.id","2026-08-19 20:08:18",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add Reihantt6\u002Fsns-myagent --skill motion-graphics",{"tool":39,"command":40},"bunx","bunx skills add Reihantt6\u002Fsns-myagent --skill motion-graphics",{"tool":42,"command":43},"pnpm","pnpm dlx skills add Reihantt6\u002Fsns-myagent --skill motion-graphics","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002FSKILL.md",true,[47,53,57,61,65,69,73,77,81,85,91,95,99,103,107,111,115,119,123,127,131,135],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"agents\u002Fbuilder.md","REFERENCE","参考文档",3394,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fagents\u002Fbuilder.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"agents\u002Fdirector.md",5756,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fagents\u002Fdirector.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"agents\u002Ffinalize.md",1071,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fagents\u002Ffinalize.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"catalog-map.md",10595,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcatalog-map.md",{"path":66,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":67,"sourceHtmlUrl":68},"categories\u002Fasset-fusion\u002Fmodule.md",3968,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fasset-fusion\u002Fmodule.md",{"path":70,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":71,"sourceHtmlUrl":72},"categories\u002Fcharts\u002Fmodule.md",1611,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fcharts\u002Fmodule.md",{"path":74,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":75,"sourceHtmlUrl":76},"categories\u002Fkinetic-type\u002Fmodule.md",1607,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fkinetic-type\u002Fmodule.md",{"path":78,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":79,"sourceHtmlUrl":80},"categories\u002Flogo-reveal\u002Fmodule.md",1131,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Flogo-reveal\u002Fmodule.md",{"path":82,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":83,"sourceHtmlUrl":84},"categories\u002Flower-thirds\u002Fmodule.md",1081,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Flower-thirds\u002Fmodule.md",{"path":86,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":89,"sourceHtmlUrl":90},"categories\u002Fmaps\u002Fbake-basemap.mjs","SCRIPT","脚本",14726,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fmaps\u002Fbake-basemap.mjs",{"path":92,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":93,"sourceHtmlUrl":94},"categories\u002Fmaps\u002Fmodule.md",9587,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fmaps\u002Fmodule.md",{"path":96,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":97,"sourceHtmlUrl":98},"categories\u002Fnews\u002Fmodule.md",5404,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fnews\u002Fmodule.md",{"path":100,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":101,"sourceHtmlUrl":102},"categories\u002Fstat\u002Fmodule.md",1200,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fstat\u002Fmodule.md",{"path":104,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":105,"sourceHtmlUrl":106},"categories\u002Ftweet\u002Fmodule.md",1015,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Ftweet\u002Fmodule.md",{"path":108,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":109,"sourceHtmlUrl":110},"categories\u002Fwebpage\u002Fmodule.md",2538,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fcategories\u002Fwebpage\u002Fmodule.md",{"path":112,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":113,"sourceHtmlUrl":114},"grounding\u002Flocate.mjs",8236,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fgrounding\u002Flocate.mjs",{"path":116,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":117,"sourceHtmlUrl":118},"grounding\u002FPROTOCOL.md",3108,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fgrounding\u002FPROTOCOL.md",{"path":120,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":121,"sourceHtmlUrl":122},"phases\u002Fsource\u002Fguide.md",1962,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fphases\u002Fsource\u002Fguide.md",{"path":124,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":125,"sourceHtmlUrl":126},"references\u002Fbuilder-contract.md",2818,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Freferences\u002Fbuilder-contract.md",{"path":128,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":129,"sourceHtmlUrl":130},"references\u002Fmotion-vocabulary.md",3566,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Freferences\u002Fmotion-vocabulary.md",{"path":132,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":133,"sourceHtmlUrl":134},"references\u002Fshot-plan-ir.md",2616,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Freferences\u002Fshot-plan-ir.md",{"path":136,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":137,"sourceHtmlUrl":138},"samples\u002Fasset-fusion\u002F_ref-circle-highlight.html",10674,"https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fmotion-graphics\u002Fsamples\u002Fasset-fusion\u002F_ref-circle-highlight.html",[140,147,154,161,168,175],{"skillNo":141,"name":142,"nameZh":6,"description":143,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":144,"categories":145,"keywords":146,"publishedAt":19},"01M77XDCYZ407XYMF2T4P6F7VT","youtube-transcript","Extract transcripts from YouTube videos. Use when the user asks for a transcript, subtitles, or captions of a YouTube video and provides a YouTube URL (youtube.com\u002Fwatch?v=, youtu.be\u002F, or similar). Supports output with or without timestamps.","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fyoutube-transcript\u002FSKILL.md",[],[],{"skillNo":148,"name":149,"nameZh":6,"description":150,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":151,"categories":152,"keywords":153,"publishedAt":19},"01M77XD7PBC98J6MXGJJ2TJ8DX","youtube-thumbnail-design","YouTube thumbnail design with specific dimensions, contrast rules, and mobile preview optimization. Covers safe zones, text placement, face expression psychology, and A\u002FB testing. Use for: YouTube thumbnails, video cover images, click-through optimization. Triggers: youtube thumbnail, thumbnail design, video thumbnail, click through rate, ctr optimization, youtube cover, video cover image, thumbnail maker, thumbnail tips, youtube design, video preview image","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fyoutube-thumbnail-design\u002FSKILL.md",[],[],{"skillNo":155,"name":156,"nameZh":6,"description":157,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":158,"categories":159,"keywords":160,"publishedAt":19},"01M77XD2541C8HXJ6RHYZP4NGJ","youtube-short-maker","Make YouTube Shorts that hold attention with Pexo — vertical 9:16, fast-paced, hook in the first 3 seconds. Describe your topic and Pexo writes the script, generates the shots, picks the models, and assembles a finished Short with music and captions. Use for YouTube Shorts: \"youtube short\", \"make a youtube short\", \"shorts video\", \"short-form video\". NOT for long-form or landscape video.","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fyoutube-short-maker\u002FSKILL.md",[],[],{"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":167,"publishedAt":19},"01M77XCXKVSYK7KTZ3PQBVVGFD","xcode-project-setup","Safely modifies Xcode projects (.pbxproj) to add Swift Packages and link files. Use this skill whenever an iOS project needs dependencies installed (e.g. Firebase, Alamofire).","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fxcode-project-setup\u002FSKILL.md",[],[],{"skillNo":169,"name":170,"nameZh":6,"description":171,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":172,"categories":173,"keywords":174,"publishedAt":19},"01M77XCR1DV8ZQK2NHWKVY4EWX","x402","Set up Browser Use Cloud payments with x402 — pay per request from a crypto wallet (USDC on Base mainnet), no signup or API key. Two setups it works out up front — \"just use it\" (set up a wallet so you or Claude Code can run cloud browser tasks paid from the wallet — Claude writes and runs throwaway scripts, nothing touches your codebase) or \"build it in\" (install the SDK and write the key + code into your project). Walks through wallet setup, funding, .env, and a ~$1 test run. Use when the user asks about x402, pay-per-use, USDC payments, or wants Browser Use Cloud without an API key. For the free-tier signup (reverse-CAPTCHA → API key), use `browser-use cloud signup` or the `cloud` skill instead.","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fx402\u002FSKILL.md",[],[],{"skillNo":176,"name":177,"nameZh":6,"description":178,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":179,"categories":180,"keywords":181,"publishedAt":19},"01M77XCMD3R3ZBZQQSW6E0793B","x-twitter-scraper","Build GitHub Copilot workflows with Xquik X API SDKs, REST endpoints, hosted Apify Actor runs, MCP tools, TweetClaw OpenClaw plugin installs, signed webhooks, tweet search, user lookup, follower exports, media actions, and agent automation.","https:\u002F\u002Fgithub.com\u002FReihantt6\u002Fsns-myagent\u002Fblob\u002Fmain\u002Fagent\u002Fskills\u002Fx-twitter-scraper\u002FSKILL.md",[],[]]