[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6FPJTMT7P8QWCPJJ6NC3GR4":3,"skill-same:01M6FPJTMT7P8QWCPJJ6NC3GR4":71},{"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":16,"sourceHtmlUrl":17,"categories":18,"keywords":19,"publishedAt":39,"linkOnly":40,"markdownContent":41,"markdownBody":42,"frontMatter":43,"refName":44,"commitSha":45,"skillDirectory":46,"forksCount":14,"repositoryHtmlUrl":47,"repositoryHomepageUrl":48,"ownerHtmlUrl":49,"ownerBlogUrl":50,"ownerDescription":51,"updatedAt":52,"install":53,"resources":66},"01M6FPJTMT7P8QWCPJJ6NC3GR4","adhd",null,"Parallel divergent ideation for coding agents. Spawns N isolated branches under different cognitive frames (regulator, biology, speedrunner, 10-year-old, $0 budget), scores, clusters, prunes traps, and deepens top survivors. Use on \u002Fadhd, \"ADHD mode\", brainstorm\u002Fideate intents, or open-ended design, architecture, naming, API\u002FSDK surface, and fuzzy-debugging decisions. Skip for syntax, lookups, bugs with known root cause, or closed phrasing (\"quick\", \"standard\", \"canonical\", \"textbook\"). Full pre-flight gate is in the skill body.","stantheman0128\u002Fstan-claude-harness","stantheman0128","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F79733497?v=4","01M6CVVRE7QYYJKA18AVBH8AZ8","Po-Han (Stan) Shih · 施博瀚","01M6CVVRJBBWCVD9RKSS251BJG",0,"MIT","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fadhd\u002FLICENSE","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fadhd\u002FSKILL.md",[],[5,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38],"brainstorm","naming","quick","with","biology","branches","ideation","use","decisions","design","parallel","open","intents","api","Budget","debugging","architecture","sdk","coding agents","2026-11-10 10:12:30",false,"---\nname: adhd\ndescription: Parallel divergent ideation for coding agents. Spawns N isolated branches under different cognitive frames (regulator, biology, speedrunner, 10-year-old, $0 budget), scores, clusters, prunes traps, and deepens top survivors. Use on \u002Fadhd, \"ADHD mode\", brainstorm\u002Fideate intents, or open-ended design, architecture, naming, API\u002FSDK surface, and fuzzy-debugging decisions. Skip for syntax, lookups, bugs with known root cause, or closed phrasing (\"quick\", \"standard\", \"canonical\", \"textbook\"). Full pre-flight gate is in the skill body.\nlicense: MIT\n---\n\n# ADHD\n\nStop picking the textbook answer. The first three answers the model would\ngive are the answers a senior engineer would give in thirty seconds.\nCorrect. Forgettable. The interesting answers live past number three, in\nthe awkward middle nobody walks into. This skill makes the model walk\nthere.\n\n## Pre-flight (run before Phase 1)\n\nThis skill is expensive. About 10 Agent calls, 30 to 90 seconds wall clock,\n5 to 10x a single answer. Do not pay that cost when a direct answer is\nbetter. Run this gate before Phase 1.\n\n**Step 1. Explicit invocation check.**\n\nIf the user typed `\u002Fadhd` or explicitly asked for ADHD mode, \"use the\nadhd skill\", or \"run ADHD on this\", **SKIP the rest of this section and go\nstraight to Phase 1**. The user opted in. Do not second-guess.\n\n**Step 2. Self-judge (only if Step 1 did not match).**\n\nAsk yourself three questions. If the answer to any is no, ABORT.\n\n1. **Open-ended?** Would a senior engineer give multiple viable answers\n   here, or is there one canonical answer? If canonical, abort.\n2. **High-stakes?** Is the cost of the obvious answer being wrong actually\n   high? Architecture decisions, public API surfaces, naming a real\n   product, fuzzy bugs with no known root cause, schema design = yes.\n   Side project at 11pm = no.\n3. **Open phrasing?** Did the user avoid words like \"quick\", \"standard\",\n   \"canonical\", \"textbook\", \"just\", \"one-line\"? If they used any of those,\n   they want the direct answer. Abort.\n\nIf all three checks pass, proceed to Phase 1.\n\nIf any fails, ABORT and answer the question directly. Optionally append\none sentence: *\"If you want a wider exploration under parallel cognitive\nframes with explicit trap detection, run `\u002Fadhd \u003Cyour problem>`.\"*\n\n## The loop\n\nTwo strict phases. Mixing them kills idea quality, because the critic\nstrangles the generator.\n\n### Phase 1 — Diverge (no critic)\n\nFor the problem P:\n\n1. Pick 5 cognitive frames from the table below. Bias toward engineering\n   tags when the problem is code-shaped. Always include at least one wild\n   frame to keep range.\n\n2. Spawn 5 **parallel** Agent\u002FTask tool calls. One per frame. Each Agent\n   gets only:\n   - the problem P\n   - any context the user provided\n   - the chosen frame's vantage prompt\n   - a system instruction that forbids evaluation\n\n   The exact instruction to give each Agent:\n\n   > You are in DIVERGENT mode. You are a generator, not a critic.\n   > Generate 6 short distinct ideas under this frame. Each idea is one\n   > phrase or one sentence. Do not evaluate. Do not rank. Do not hedge.\n   > The first three obvious answers everyone would give are banned.\n   > Push past them into the awkward middle.\n   > Output a JSON array only. No prose before or after.\n   > `[{\"text\": \"...\", \"rationale\": \"...\"}, ...]`\n\n3. **Critical invariant.** The Agent calls must be parallel and isolated.\n   Do NOT serialize them. Do NOT pass one branch's output as context to\n   another. Branches that see each other anchor each other and the whole\n   method collapses to a wider single thought.\n\n### Phase 2 — Focus (critic on)\n\nAfter all branches return:\n\n1. **Score.** Rate each idea on three axes 0 to 10: novelty (distance from\n   the obvious default), viability (could it actually ship), fit (does it\n   address the stated problem). For any idea that looks attractive but is\n   a trap (hidden cost, false economy, will not scale, premature\n   abstraction), flag it with a one-line reason.\n\n2. **Cluster.** Group ideas into 3 to 6 clusters by their underlying angle,\n   not by surface keywords. Label clusters by angle: \"remove the server\n   plays\", \"cache-shaped plays\", \"batched-window plays\", \"race-multiple-\n   backends plays\".\n\n3. **Deepen the top 3.** Rank by weighted score (novelty 0.35 + viability\n   0.40 + fit 0.25), exclude traps, take top 3. For each, spawn one Agent\n   call that produces:\n   - a 4 to 8 sentence sketch of how the idea works\n   - the load-bearing risk\n   - the first concrete step a builder would take\n   - 3 to 5 child ideas (variations, hybrids, unlocks)\n\n   Deepen Agent instruction:\n\n   > You are in FOCUS mode. Take one promising idea and connect dots.\n   > Sketch how it would actually work in 4 to 8 sentences. Name the\n   > load-bearing risk. Name the first concrete step a coder would take.\n   > Then generate 3 to 5 sub-ideas that branch off (variations,\n   > combinations with other domains, things this unlocks).\n   > Output JSON only.\n\n## Frames\n\nPick 5 per run.\n\n| Frame | Vantage prompt | Tags |\n|---|---|---|\n| **hardware engineer** | You think in latency, memory layout, and physical constraints. Re-ask this as a hardware\u002Ffirmware problem. What does the bus topology, cache, timing budget tell you? | code, wild |\n| **regulator** | You audit systems for compliance and failure modes. What must be provable, traceable, or refusable here? | design, general |\n| **10-year-old** | You are a curious 10 year old who has never seen software. Describe naive but unencumbered approaches. Ignore convention. | general, wild |\n| **competitor trying to break it** | You are a hostile competitor or attacker. Generate approaches that exploit, fail, or sabotage the obvious solution. Then invert into ideas. | code, design |\n| **biology** | Transplant a mechanism from biology (immune systems, neural plasticity, cell signaling, evolution, gut flora). Force-fit it onto this engineering problem. | code, wild |\n| **logistics** | Steal mechanisms from logistics: queues, batching, just-in-time, hub-and-spoke, returns, last-mile. Apply them literally. | code, design |\n| **game design** | Approach this as a game designer. What are the loops, rewards, friction, save-states, speedrun tricks? Treat the user as a player. | design, general |\n| **markets** | Treat the problem as a market. Buyers, sellers, market-makers. What does an auction, a futures contract, a clearing house look like here? | design, wild |\n| **inversion** | Ask the OPPOSITE question. If goal is X, brainstorm how to guarantee NOT X. Then negate each answer back. | code, design, general |\n| **extreme: $0 budget, 1 hour** | No money, no team, one hour. What is the crudest version that still does the load-bearing thing? | code, general |\n| **extreme: infinite budget, 10 years** | Infinite compute, infinite engineers, a decade. What is the maximalist version? | design, wild |\n| **remove the load-bearing assumption** | Name the thing everyone treats as fixed (framework, database, request-response model, network). Imagine it is gone. What is possible? | code, design, wild |\n| **speedrunner** | You are a speedrunner. Find glitches, skips, out-of-bounds tricks, frame-perfect shortcuts. What is the abusive-but-legal path? | code, wild |\n| **ant colony** | No central planner. Many dumb agents, local rules, pheromone trails. How does the problem solve itself emergently? | code, wild |\n| **3am on-call** | You are the on-call engineer woken at 3am when this breaks. What design would let you not get paged? | code, design |\n\n### Picking frames\n\nFor code-shaped problems: pick 4 frames tagged `code` or `design`, plus 1\ntagged `wild`. For open product or strategy problems: a mix from all tags.\nVary the picks across sessions so the same problem produces different\ncandidate sets when re-run.\n\n## Output shape\n\nAfter Phase 2, render in this order. Do not collapse it into a wall of\nprose. The structure is the point.\n\n1. **Brief.** One or two lines confirming the problem and any reframe used.\n2. **Wide set.** Full pool grouped by cluster. Each cluster labeled by\n   underlying angle. Each idea is one short phrase. Show score chips like\n   `[N7 V8 F9]` next to each.\n3. **Converge.** A 2 to 4 idea shortlist. State why each is on the list.\n   Mark the non-obvious-but-viable pick explicitly with ★. List traps\n   separately, each with the one-line reason it is a trap.\n4. **Focus.** The 3 deepened branches. For each: the sketch, the load-\n   bearing risk, the first concrete step, and the child ideas.\n5. **Provocation.** One wildcard question or idea that opens a new\n   direction the user can push into if nothing landed.\n\n## Anti-patterns\n\nThese are how this skill goes wrong. Watch for them.\n\n- **Convergence disguised as divergence.** Ten minor variations of one idea\n  is not breadth. If every candidate shares the same underlying assumption,\n  you have not diverged. You have decorated.\n- **Weird-for-weird's-sake with no convergence.** A pile of 30 unsorted\n  absurdities is as useless as one safe answer. Always converge.\n- **Walls of equally-weighted prose.** Cluster, label, pull out the best.\n  Structure is half the value.\n- **Refusing to commit.** After diverging, take a position on what is\n  actually promising. \"Here are 20 ideas, you decide\" is a cop-out.\n  Generate wide, but converge with a real opinion.\n- **Skipping the isolation invariant.** If you simulate parallel branches\n  by writing them sequentially in one context, you have not done ADHD. You\n  have done a wider single thought. The Agent\u002FTask tool gives each branch a\n  fresh context. Use it.\n\n## Calibration\n\n- **How many ideas?** Scale to stakes. Quick \"name this function\" =\n  3 frames × 4 ideas. \"How should I position this product\" = 5 frames ×\n  8 ideas. Default is 5 × 6 = 30.\n- **How weird?** Read the room. Serious strategy work: flag the wild cards\n  clearly so they do not read as unserious. Open brainstorming or play:\n  let it run loose. Absurd ideas earn their place by seeding viable ones.\n- **When to stop diverging?** Stop when new candidates start repeating the\n  shape of existing ones. The space is mapped. Do not pad to hit a number.\n\n## Cost\n\n5 diverge + 1 score + 1 cluster + 3 deepen ≈ 10 Agent calls per run.\nAbout 5 to 10x a single-shot answer. Not for every keystroke. For decision\npoints where the cost of the obvious answer is high.\n\n## Companion library and CLI\n\nThere is a Node\u002FTS implementation that does the same loop with structured\nJSON parsing, score weighting, and a CLI. Use it when running outside\nClaude Code or in batch.\n\n    npm install -g adhd-agent\n    adhd \"your problem here\"\n\nCode, paper, evals, and contributing guide at\nhttps:\u002F\u002Fgithub.com\u002FUditAkhourii\u002Fadhd. The skill above gives you the same\nloop inside Claude with no install required.\n\n## Source spec\n\nThis skill operationalises a written spec on divergent ideation. The\noriginal prose is preserved in `SOURCE-SPEC.md` for reference. The\nimplementation choices made here (parallel isolated Agent calls,\nmechanical generator\u002Fcritic split, frame-based branching) follow from\nthat spec.\n","# ADHD\n\nStop picking the textbook answer. The first three answers the model would\ngive are the answers a senior engineer would give in thirty seconds.\nCorrect. Forgettable. The interesting answers live past number three, in\nthe awkward middle nobody walks into. This skill makes the model walk\nthere.\n\n## Pre-flight (run before Phase 1)\n\nThis skill is expensive. About 10 Agent calls, 30 to 90 seconds wall clock,\n5 to 10x a single answer. Do not pay that cost when a direct answer is\nbetter. Run this gate before Phase 1.\n\n**Step 1. Explicit invocation check.**\n\nIf the user typed `\u002Fadhd` or explicitly asked for ADHD mode, \"use the\nadhd skill\", or \"run ADHD on this\", **SKIP the rest of this section and go\nstraight to Phase 1**. The user opted in. Do not second-guess.\n\n**Step 2. Self-judge (only if Step 1 did not match).**\n\nAsk yourself three questions. If the answer to any is no, ABORT.\n\n1. **Open-ended?** Would a senior engineer give multiple viable answers\n   here, or is there one canonical answer? If canonical, abort.\n2. **High-stakes?** Is the cost of the obvious answer being wrong actually\n   high? Architecture decisions, public API surfaces, naming a real\n   product, fuzzy bugs with no known root cause, schema design = yes.\n   Side project at 11pm = no.\n3. **Open phrasing?** Did the user avoid words like \"quick\", \"standard\",\n   \"canonical\", \"textbook\", \"just\", \"one-line\"? If they used any of those,\n   they want the direct answer. Abort.\n\nIf all three checks pass, proceed to Phase 1.\n\nIf any fails, ABORT and answer the question directly. Optionally append\none sentence: *\"If you want a wider exploration under parallel cognitive\nframes with explicit trap detection, run `\u002Fadhd \u003Cyour problem>`.\"*\n\n## The loop\n\nTwo strict phases. Mixing them kills idea quality, because the critic\nstrangles the generator.\n\n### Phase 1 — Diverge (no critic)\n\nFor the problem P:\n\n1. Pick 5 cognitive frames from the table below. Bias toward engineering\n   tags when the problem is code-shaped. Always include at least one wild\n   frame to keep range.\n\n2. Spawn 5 **parallel** Agent\u002FTask tool calls. One per frame. Each Agent\n   gets only:\n   - the problem P\n   - any context the user provided\n   - the chosen frame's vantage prompt\n   - a system instruction that forbids evaluation\n\n   The exact instruction to give each Agent:\n\n   > You are in DIVERGENT mode. You are a generator, not a critic.\n   > Generate 6 short distinct ideas under this frame. Each idea is one\n   > phrase or one sentence. Do not evaluate. Do not rank. Do not hedge.\n   > The first three obvious answers everyone would give are banned.\n   > Push past them into the awkward middle.\n   > Output a JSON array only. No prose before or after.\n   > `[{\"text\": \"...\", \"rationale\": \"...\"}, ...]`\n\n3. **Critical invariant.** The Agent calls must be parallel and isolated.\n   Do NOT serialize them. Do NOT pass one branch's output as context to\n   another. Branches that see each other anchor each other and the whole\n   method collapses to a wider single thought.\n\n### Phase 2 — Focus (critic on)\n\nAfter all branches return:\n\n1. **Score.** Rate each idea on three axes 0 to 10: novelty (distance from\n   the obvious default), viability (could it actually ship), fit (does it\n   address the stated problem). For any idea that looks attractive but is\n   a trap (hidden cost, false economy, will not scale, premature\n   abstraction), flag it with a one-line reason.\n\n2. **Cluster.** Group ideas into 3 to 6 clusters by their underlying angle,\n   not by surface keywords. Label clusters by angle: \"remove the server\n   plays\", \"cache-shaped plays\", \"batched-window plays\", \"race-multiple-\n   backends plays\".\n\n3. **Deepen the top 3.** Rank by weighted score (novelty 0.35 + viability\n   0.40 + fit 0.25), exclude traps, take top 3. For each, spawn one Agent\n   call that produces:\n   - a 4 to 8 sentence sketch of how the idea works\n   - the load-bearing risk\n   - the first concrete step a builder would take\n   - 3 to 5 child ideas (variations, hybrids, unlocks)\n\n   Deepen Agent instruction:\n\n   > You are in FOCUS mode. Take one promising idea and connect dots.\n   > Sketch how it would actually work in 4 to 8 sentences. Name the\n   > load-bearing risk. Name the first concrete step a coder would take.\n   > Then generate 3 to 5 sub-ideas that branch off (variations,\n   > combinations with other domains, things this unlocks).\n   > Output JSON only.\n\n## Frames\n\nPick 5 per run.\n\n| Frame | Vantage prompt | Tags |\n|---|---|---|\n| **hardware engineer** | You think in latency, memory layout, and physical constraints. Re-ask this as a hardware\u002Ffirmware problem. What does the bus topology, cache, timing budget tell you? | code, wild |\n| **regulator** | You audit systems for compliance and failure modes. What must be provable, traceable, or refusable here? | design, general |\n| **10-year-old** | You are a curious 10 year old who has never seen software. Describe naive but unencumbered approaches. Ignore convention. | general, wild |\n| **competitor trying to break it** | You are a hostile competitor or attacker. Generate approaches that exploit, fail, or sabotage the obvious solution. Then invert into ideas. | code, design |\n| **biology** | Transplant a mechanism from biology (immune systems, neural plasticity, cell signaling, evolution, gut flora). Force-fit it onto this engineering problem. | code, wild |\n| **logistics** | Steal mechanisms from logistics: queues, batching, just-in-time, hub-and-spoke, returns, last-mile. Apply them literally. | code, design |\n| **game design** | Approach this as a game designer. What are the loops, rewards, friction, save-states, speedrun tricks? Treat the user as a player. | design, general |\n| **markets** | Treat the problem as a market. Buyers, sellers, market-makers. What does an auction, a futures contract, a clearing house look like here? | design, wild |\n| **inversion** | Ask the OPPOSITE question. If goal is X, brainstorm how to guarantee NOT X. Then negate each answer back. | code, design, general |\n| **extreme: $0 budget, 1 hour** | No money, no team, one hour. What is the crudest version that still does the load-bearing thing? | code, general |\n| **extreme: infinite budget, 10 years** | Infinite compute, infinite engineers, a decade. What is the maximalist version? | design, wild |\n| **remove the load-bearing assumption** | Name the thing everyone treats as fixed (framework, database, request-response model, network). Imagine it is gone. What is possible? | code, design, wild |\n| **speedrunner** | You are a speedrunner. Find glitches, skips, out-of-bounds tricks, frame-perfect shortcuts. What is the abusive-but-legal path? | code, wild |\n| **ant colony** | No central planner. Many dumb agents, local rules, pheromone trails. How does the problem solve itself emergently? | code, wild |\n| **3am on-call** | You are the on-call engineer woken at 3am when this breaks. What design would let you not get paged? | code, design |\n\n### Picking frames\n\nFor code-shaped problems: pick 4 frames tagged `code` or `design`, plus 1\ntagged `wild`. For open product or strategy problems: a mix from all tags.\nVary the picks across sessions so the same problem produces different\ncandidate sets when re-run.\n\n## Output shape\n\nAfter Phase 2, render in this order. Do not collapse it into a wall of\nprose. The structure is the point.\n\n1. **Brief.** One or two lines confirming the problem and any reframe used.\n2. **Wide set.** Full pool grouped by cluster. Each cluster labeled by\n   underlying angle. Each idea is one short phrase. Show score chips like\n   `[N7 V8 F9]` next to each.\n3. **Converge.** A 2 to 4 idea shortlist. State why each is on the list.\n   Mark the non-obvious-but-viable pick explicitly with ★. List traps\n   separately, each with the one-line reason it is a trap.\n4. **Focus.** The 3 deepened branches. For each: the sketch, the load-\n   bearing risk, the first concrete step, and the child ideas.\n5. **Provocation.** One wildcard question or idea that opens a new\n   direction the user can push into if nothing landed.\n\n## Anti-patterns\n\nThese are how this skill goes wrong. Watch for them.\n\n- **Convergence disguised as divergence.** Ten minor variations of one idea\n  is not breadth. If every candidate shares the same underlying assumption,\n  you have not diverged. You have decorated.\n- **Weird-for-weird's-sake with no convergence.** A pile of 30 unsorted\n  absurdities is as useless as one safe answer. Always converge.\n- **Walls of equally-weighted prose.** Cluster, label, pull out the best.\n  Structure is half the value.\n- **Refusing to commit.** After diverging, take a position on what is\n  actually promising. \"Here are 20 ideas, you decide\" is a cop-out.\n  Generate wide, but converge with a real opinion.\n- **Skipping the isolation invariant.** If you simulate parallel branches\n  by writing them sequentially in one context, you have not done ADHD. You\n  have done a wider single thought. The Agent\u002FTask tool gives each branch a\n  fresh context. Use it.\n\n## Calibration\n\n- **How many ideas?** Scale to stakes. Quick \"name this function\" =\n  3 frames × 4 ideas. \"How should I position this product\" = 5 frames ×\n  8 ideas. Default is 5 × 6 = 30.\n- **How weird?** Read the room. Serious strategy work: flag the wild cards\n  clearly so they do not read as unserious. Open brainstorming or play:\n  let it run loose. Absurd ideas earn their place by seeding viable ones.\n- **When to stop diverging?** Stop when new candidates start repeating the\n  shape of existing ones. The space is mapped. Do not pad to hit a number.\n\n## Cost\n\n5 diverge + 1 score + 1 cluster + 3 deepen ≈ 10 Agent calls per run.\nAbout 5 to 10x a single-shot answer. Not for every keystroke. For decision\npoints where the cost of the obvious answer is high.\n\n## Companion library and CLI\n\nThere is a Node\u002FTS implementation that does the same loop with structured\nJSON parsing, score weighting, and a CLI. Use it when running outside\nClaude Code or in batch.\n\n    npm install -g adhd-agent\n    adhd \"your problem here\"\n\nCode, paper, evals, and contributing guide at\nhttps:\u002F\u002Fgithub.com\u002FUditAkhourii\u002Fadhd. The skill above gives you the same\nloop inside Claude with no install required.\n\n## Source spec\n\nThis skill operationalises a written spec on divergent ideation. The\noriginal prose is preserved in `SOURCE-SPEC.md` for reference. The\nimplementation choices made here (parallel isolated Agent calls,\nmechanical generator\u002Fcritic split, frame-based branching) follow from\nthat spec.\n",{"name":5,"license":15,"description":7},"master","5a2ab4d2754b223fde30b452e59368798751168c","skills\u002Fadhd","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness","https:\u002F\u002Fstan-shih.com\u002Fsoftware#stan-claude-harness","https:\u002F\u002Fgithub.com\u002Fstantheman0128","https:\u002F\u002Fstan-shih.com\u002F","AI product developer building AI agents, Android apps, and developer tools.","2026-11-14 17:22:52",{"source":8,"skillName":5,"commands":54,"prompt":64,"zipAvailable":65},[55,58,61],{"tool":56,"command":57},"npx","npx skills add stantheman0128\u002Fstan-claude-harness --skill adhd",{"tool":59,"command":60},"bunx","bunx skills add stantheman0128\u002Fstan-claude-harness --skill adhd",{"tool":62,"command":63},"pnpm","pnpm dlx skills add stantheman0128\u002Fstan-claude-harness --skill adhd","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fadhd\u002FSKILL.md",true,[67],{"path":68,"resourceType":68,"resourceTypeDesc":69,"sizeBytes":70,"sourceHtmlUrl":16},"LICENSE","许可证",1074,[72,115,143],{"skillNo":73,"name":74,"nameZh":6,"description":75,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":76,"licenseFileUrl":77,"sourceHtmlUrl":78,"categories":79,"keywords":96,"publishedAt":114},"01M6FPTG8GGXS2Q7Y66ZWQPR1J","xlsx","Use this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like \"the xlsx in my downloads\") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fsynced\u002Fe7b46bc2-21f0-423a-868e-95e59f5d567a_17945918-9d6c-416c-81c9-dacad4eafd1a\u002Fxlsx\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fsynced\u002Fe7b46bc2-21f0-423a-868e-95e59f5d567a_17945918-9d6c-416c-81c9-dacad4eafd1a\u002Fxlsx\u002FSKILL.md",[80,85,88,91],{"categoryNo":81,"categoryName":82,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000002","文档处理","SCENARIO","使用场景",{"categoryNo":86,"categoryName":87,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000004","电子表格",{"categoryNo":89,"categoryName":90,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000010","后端与接口",{"categoryNo":92,"categoryName":93,"taxonomyType":94,"taxonomyTypeDesc":95},"01JBSEEDCAT000000000000034","行政与办公人员","PROFESSION","适用职业",[74,97,98,99,100,101,102,103,104,31,105,106,107,33,108,109,110,111,112,113],"create","sources","spreadsheet","pipeline","time","python script","fix","html report","files","database","done","references","Charting","read","standalone","when","data sources","2026-11-11 15:11:14",{"skillNo":116,"name":117,"nameZh":6,"description":118,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":119,"categories":120,"keywords":124,"publishedAt":39},"01M6FPVHJPHA7176860RTHDA9C","video-lens","Summarize YouTube URLs or video IDs from transcripts into HTML reports with key points and timestamps. Use for video summaries, talk notes, or breakdowns; supports transcript-language selection.","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fvideo-lens\u002FSKILL.md",[121],{"categoryNo":122,"categoryName":123,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000016","图像与多媒体",[125,126,127,128,129,130,131,132,133,134,135,33,136,137,138,139,140,141,142,110],"Video","lens","video lens","dlp","Captions","ids","local","selection","only","timestamps","Bash","mlx","transcript","reports","whisper","notes","apple","youtube",{"skillNo":144,"name":145,"nameZh":6,"description":146,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":147,"categories":148,"keywords":150,"publishedAt":39},"01M6FPV8ZERVH7W9RASR0N1V5G","video-lens-gallery","Browse, search, open, or maintain saved video-lens reports. Use for the video-summary library, rebuilding its index, or requested metadata backfills.","https:\u002F\u002Fgithub.com\u002Fstantheman0128\u002Fstan-claude-harness\u002Fblob\u002Fmaster\u002Fskills\u002Fvideo-lens-gallery\u002FSKILL.md",[149],{"categoryNo":122,"categoryName":123,"taxonomyType":83,"taxonomyTypeDesc":84},[151,152,126,125,153,27,154,135,155,31,156,138,157,158,110,159],"video lens gallery","gallery","maintain","library","metadata","index","browse","search","summary"]