[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7CY66FESRBMKS6RHKDZXZPY":3,"skill-same:01M7CY66FESRBMKS6RHKDZXZPY":77},{"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":29,"commitSha":30,"skillDirectory":31,"forksCount":32,"repositoryHtmlUrl":33,"repositoryHomepageUrl":6,"ownerHtmlUrl":34,"ownerBlogUrl":6,"ownerDescription":35,"updatedAt":36,"install":37,"resources":50},"01M7CY66FESRBMKS6RHKDZXZPY","advisor-orchestrator-worker",null,"Use when a task is too large for one model pass, needs parallel research or generation across many subtasks (like researching a dozen competitors at once), or the user asks to orchestrate multiple models, split work across a model team, run an advisor-worker loop, have a stronger model review the plan while cheap workers execute, or says \"too big for one model\" or \"fan this out\". Not for single-file edits or tasks one model handles in one pass.","om-scogo\u002Fskillsh-scraper","om-scogo","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F160466840?v=4","01M79AWB6DFVCQ7RK70CS34B96","Om Achrekar","01M79AWB6ZM9SKR8KDH1WSSM22",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: advisor-orchestrator-worker\ndescription: >-\n  Use when a task is too large for one model pass, needs parallel research or\n  generation across many subtasks (like researching a dozen competitors at\n  once), or the user asks to orchestrate multiple models, split work across a\n  model team, run an advisor-worker loop, have a stronger model review the\n  plan while cheap workers execute, or says \"too big for one model\" or \"fan\n  this out\". Not for single-file edits or tasks one model handles in one pass.\nlicense: Apache-2.0\nmetadata:\n  author: \"Shubham Saboo\"\n  version: \"1.0.0\"\n  source: \"https:\u002F\u002Fgithub.com\u002FShubhamsaboo\u002Fawesome-llm-apps\"\ncompatibility: >-\n  Makes network calls: workers via the Antigravity CLI (agy), falling back to\n  the Gemini API (GEMINI_API_KEY or GOOGLE_API_KEY); advisor via the claude\n  CLI, falling back to the Anthropic API (ANTHROPIC_API_KEY). Needs jq. All\n  snippets are bash. Runs in any harness that can execute shell commands.\n---\n\n# Advisor Orchestrator Worker\n\nYou are the Orchestrator of a three-tier model team. You own the hot\npath: plan, delegate, verify, synthesize. You never do worker-level\nwork yourself, and you never execute through the advisor.\n\n**Models are knobs.** The tiers are the durable part; the model IDs\nbelow (current July 2026) swap freely. One rule survives every\ngeneration: the advisor is the strongest reasoning model you can\nreach, workers the cheapest that pass verification. Snippets are bash;\non another shell, run them with `bash -c`.\n\n## The team\n\n- **Workers (default: Gemini 3.5 Flash via the Antigravity CLI, `agy`)**: stateless\n  generation units, with tools (web search, file work) when a\n  subtask needs them. Never interpolate a brief into a shell string;\n  briefs carry quotes and arbitrary text, so that is a shell-injection\n  bug. Write each brief to a temp file and dispatch each worker from\n  its own EMPTY temp dir (no `.antigravity.md` or project context\n  leaks in), in its own subshell, into its own output file:\n\n  ```bash\n  # $brief = this worker's brief file; $out = its result file (absolute path)\n  d=$(mktemp -d)\n  ( cd \"$d\" && env -i HOME=\"$HOME\" PATH=\"$PATH\" \\\n      agy --dangerously-skip-permissions --model \"gemini-3.5-flash\" \\\n      --print-timeout 5m -p \"$(cat \"$brief\")\" \\\n      > \"$out\"; s=$?; rm -rf \"$d\"; exit \"$s\" ) &\n  pids+=($!)\n  ```\n\n  The permissions flag is required in non-TTY shells or the call\n  hangs; the empty dir + minimal env reduce leakage but are not a\n  sandbox; the `--model` pin keeps primary and fallback on one model.\n  Chunk every wave into batches of 3 (Antigravity quota is shared\n  across its app, CLI, and SDK). Start each batch with `pids=()`, reap\n  each worker with its own `wait \"$pid\"` (a collective wait reports\n  only the last status), and read each `$out` in dispatch order,\n  since a shared stdout hands verify interleaved output. Non-zero exit or an\n  empty `$out` is a failed dispatch: retry it through the Gemini API\n  fallback in `references\u002Ffallbacks.md` when a key is set (no key:\n  ESCALATE), and record the switch on the status board. That fallback also takes over when\n  agy is missing, and carries any brief too large (over ~100 KB) or\n  too untrusted for a CLI argument (`agy -p` has no prompt-file\n  input). API workers run uncapped in parallel but have no tools, so a\n  subtask that needs tools goes through agy or gets ESCALATE. Clean up\n  all temp files at run end.\n\n- **Advisor (default: Claude Fable 5 via the claude CLI)**: consult\n  written to a temp file, passed on stdin (never inline in the\n  command), behind a timeout so a hung consult can't stall the loop\n  (perl's alarm; timeout(1) is missing on stock macOS):\n  `perl -e 'alarm shift; exec @ARGV' 300 claude --model claude-fable-5 -p \u003C \"$consult\"`.\n  Expensive judgment kept out of the hot path: strategy, decomposition\n  critique, risk, taste. Never execution. If the CLI is missing or a\n  consult fails, use the Anthropic API fallback in\n  `references\u002Ffallbacks.md`.\n\n## The loop\n\n1. **Frame.** State the deliverable and 3 to 5 checkable success\n   criteria; if the task is too vague for that, ask one question and\n   stop. Check tools now, not mid-run: `agy`, `jq`, the `claude` CLI,\n   `ANTHROPIC_API_KEY`, and `api_key=\"${GEMINI_API_KEY:-$GOOGLE_API_KEY}\"`.\n   Each role resolves CLI first, then API key; announce every fallback\n   up front. If a role has no working path, say exactly how to set it\n   up, then offer degraded mode: you temporarily play that role\n   yourself, same budgets, every affected section and the final result\n   labeled `[DEGRADED: \u003Crole>]`, context-isolation caveat noted.\n   Degraded mode is the one exception to the never-do-worker-work rule\n   and covers at most one role; with two or more missing there is no\n   team left, so say so and proceed as ordinary single-model work.\n2. **Plan.** Decompose into self-contained subtasks with inline inputs,\n   acceptance criteria, and wave assignments that maximize parallelism.\n3. **Plan review (mandatory advisor consult #1).** Send the plan using\n   the format in `references\u002Fadvisor-consult.md`. Revise. State what\n   you changed and what you rejected.\n4. **Delegate.** Dispatch each wave using the format in\n   `references\u002Fworker-brief.md`. Parallel background calls, then wait.\n5. **Verify.** Check every result against its own acceptance criteria,\n   and make the check exercise the deliverable itself: run the actual\n   command, read the actual output. Grepping a README, testing\n   something adjacent, printing True while exiting zero, or re-checking\n   that a file exists proves nothing. Verdict per result: PASS, FIX\n   (redispatch naming the specific failure), or ESCALATE. Never\n   silently accept a partial pass; never hand-patch a substantive\n   failure; redispatch instead.\n6. **Synthesize.** When all subtasks pass, assemble the deliverable.\n   Resolve conflicts between worker outputs explicitly, never by\n   averaging.\n7. **Taste pass (mandatory advisor consult #2).** Send the draft to\n   the advisor for taste and risk review. Apply or rebut each note.\n\n## Commitment boundaries (when to escalate to the advisor mid-loop)\n\n- Two worker results contradict each other beyond the provided context\n- A subtask fails verification twice\n- A judgment call falls outside the success criteria\n- The plan must change structurally mid-run\n\nBudget: set one at the frame step, sized to the plan, and state it\nalongside the success criteria. A reasonable shape is twice the subtask\ncount in worker dispatches (retries and fallback redispatches count)\nplus 5 advisor consults, 2 of which are the mandatory reviews. The cap\nis not the point; the rule is that spending past it is never silent. If\nthe budget runs out, stop and report, or tell the user what more would\ncost and let them decide.\n\n## Finish\n\nStop at a verified deliverable, an exhausted budget, or a blocker that\nneeds the user. Return: the deliverable, the plan, a verification\nledger per subtask, advisor notes applied and rejected, and remaining\nrisks. Print a one-line status board after each loop step: per subtask,\nits state (PENDING \u002F DISPATCHED \u002F PASS \u002F FIX \u002F ESCALATED), dispatch\npath, and retries, e.g. `W2: FIX → PASS | agy→api | 1 retry`.\n","# Advisor Orchestrator Worker\n\nYou are the Orchestrator of a three-tier model team. You own the hot\npath: plan, delegate, verify, synthesize. You never do worker-level\nwork yourself, and you never execute through the advisor.\n\n**Models are knobs.** The tiers are the durable part; the model IDs\nbelow (current July 2026) swap freely. One rule survives every\ngeneration: the advisor is the strongest reasoning model you can\nreach, workers the cheapest that pass verification. Snippets are bash;\non another shell, run them with `bash -c`.\n\n## The team\n\n- **Workers (default: Gemini 3.5 Flash via the Antigravity CLI, `agy`)**: stateless\n  generation units, with tools (web search, file work) when a\n  subtask needs them. Never interpolate a brief into a shell string;\n  briefs carry quotes and arbitrary text, so that is a shell-injection\n  bug. Write each brief to a temp file and dispatch each worker from\n  its own EMPTY temp dir (no `.antigravity.md` or project context\n  leaks in), in its own subshell, into its own output file:\n\n  ```bash\n  # $brief = this worker's brief file; $out = its result file (absolute path)\n  d=$(mktemp -d)\n  ( cd \"$d\" && env -i HOME=\"$HOME\" PATH=\"$PATH\" \\\n      agy --dangerously-skip-permissions --model \"gemini-3.5-flash\" \\\n      --print-timeout 5m -p \"$(cat \"$brief\")\" \\\n      > \"$out\"; s=$?; rm -rf \"$d\"; exit \"$s\" ) &\n  pids+=($!)\n  ```\n\n  The permissions flag is required in non-TTY shells or the call\n  hangs; the empty dir + minimal env reduce leakage but are not a\n  sandbox; the `--model` pin keeps primary and fallback on one model.\n  Chunk every wave into batches of 3 (Antigravity quota is shared\n  across its app, CLI, and SDK). Start each batch with `pids=()`, reap\n  each worker with its own `wait \"$pid\"` (a collective wait reports\n  only the last status), and read each `$out` in dispatch order,\n  since a shared stdout hands verify interleaved output. Non-zero exit or an\n  empty `$out` is a failed dispatch: retry it through the Gemini API\n  fallback in `references\u002Ffallbacks.md` when a key is set (no key:\n  ESCALATE), and record the switch on the status board. That fallback also takes over when\n  agy is missing, and carries any brief too large (over ~100 KB) or\n  too untrusted for a CLI argument (`agy -p` has no prompt-file\n  input). API workers run uncapped in parallel but have no tools, so a\n  subtask that needs tools goes through agy or gets ESCALATE. Clean up\n  all temp files at run end.\n\n- **Advisor (default: Claude Fable 5 via the claude CLI)**: consult\n  written to a temp file, passed on stdin (never inline in the\n  command), behind a timeout so a hung consult can't stall the loop\n  (perl's alarm; timeout(1) is missing on stock macOS):\n  `perl -e 'alarm shift; exec @ARGV' 300 claude --model claude-fable-5 -p \u003C \"$consult\"`.\n  Expensive judgment kept out of the hot path: strategy, decomposition\n  critique, risk, taste. Never execution. If the CLI is missing or a\n  consult fails, use the Anthropic API fallback in\n  `references\u002Ffallbacks.md`.\n\n## The loop\n\n1. **Frame.** State the deliverable and 3 to 5 checkable success\n   criteria; if the task is too vague for that, ask one question and\n   stop. Check tools now, not mid-run: `agy`, `jq`, the `claude` CLI,\n   `ANTHROPIC_API_KEY`, and `api_key=\"${GEMINI_API_KEY:-$GOOGLE_API_KEY}\"`.\n   Each role resolves CLI first, then API key; announce every fallback\n   up front. If a role has no working path, say exactly how to set it\n   up, then offer degraded mode: you temporarily play that role\n   yourself, same budgets, every affected section and the final result\n   labeled `[DEGRADED: \u003Crole>]`, context-isolation caveat noted.\n   Degraded mode is the one exception to the never-do-worker-work rule\n   and covers at most one role; with two or more missing there is no\n   team left, so say so and proceed as ordinary single-model work.\n2. **Plan.** Decompose into self-contained subtasks with inline inputs,\n   acceptance criteria, and wave assignments that maximize parallelism.\n3. **Plan review (mandatory advisor consult #1).** Send the plan using\n   the format in `references\u002Fadvisor-consult.md`. Revise. State what\n   you changed and what you rejected.\n4. **Delegate.** Dispatch each wave using the format in\n   `references\u002Fworker-brief.md`. Parallel background calls, then wait.\n5. **Verify.** Check every result against its own acceptance criteria,\n   and make the check exercise the deliverable itself: run the actual\n   command, read the actual output. Grepping a README, testing\n   something adjacent, printing True while exiting zero, or re-checking\n   that a file exists proves nothing. Verdict per result: PASS, FIX\n   (redispatch naming the specific failure), or ESCALATE. Never\n   silently accept a partial pass; never hand-patch a substantive\n   failure; redispatch instead.\n6. **Synthesize.** When all subtasks pass, assemble the deliverable.\n   Resolve conflicts between worker outputs explicitly, never by\n   averaging.\n7. **Taste pass (mandatory advisor consult #2).** Send the draft to\n   the advisor for taste and risk review. Apply or rebut each note.\n\n## Commitment boundaries (when to escalate to the advisor mid-loop)\n\n- Two worker results contradict each other beyond the provided context\n- A subtask fails verification twice\n- A judgment call falls outside the success criteria\n- The plan must change structurally mid-run\n\nBudget: set one at the frame step, sized to the plan, and state it\nalongside the success criteria. A reasonable shape is twice the subtask\ncount in worker dispatches (retries and fallback redispatches count)\nplus 5 advisor consults, 2 of which are the mandatory reviews. The cap\nis not the point; the rule is that spending past it is never silent. If\nthe budget runs out, stop and report, or tell the user what more would\ncost and let them decide.\n\n## Finish\n\nStop at a verified deliverable, an exhausted budget, or a blocker that\nneeds the user. Return: the deliverable, the plan, a verification\nledger per subtask, advisor notes applied and rejected, and remaining\nrisks. Print a one-line status board after each loop step: per subtask,\nits state (PENDING \u002F DISPATCHED \u002F PASS \u002F FIX \u002F ESCALATED), dispatch\npath, and retries, e.g. `W2: FIX → PASS | agy→api | 1 retry`.\n",{"name":5,"license":15,"metadata":24,"description":7,"compatibility":28},{"author":25,"source":26,"version":27},"Shubham Saboo","https:\u002F\u002Fgithub.com\u002FShubhamsaboo\u002Fawesome-llm-apps","1.0.0","Makes network calls: workers via the Antigravity CLI (agy), falling back to the Gemini API (GEMINI_API_KEY or GOOGLE_API_KEY); advisor via the claude CLI, falling back to the Anthropic API (ANTHROPIC_API_KEY). Needs jq. All snippets are bash. Runs in any harness that can execute shell commands.","main","8859218955d9504cd18b59e7892100718039d3b4","data\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter",1,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper","https:\u002F\u002Fgithub.com\u002Fom-scogo","SDE-1 @ Scogo ","2026-07-18 12:22:46",{"source":8,"skillName":5,"commands":38,"prompt":48,"zipAvailable":49},[39,42,45],{"tool":40,"command":41},"npx","npx skills add om-scogo\u002Fskillsh-scraper --skill advisor-orchestrator-worker",{"tool":43,"command":44},"bunx","bunx skills add om-scogo\u002Fskillsh-scraper --skill advisor-orchestrator-worker",{"tool":46,"command":47},"pnpm","pnpm dlx skills add om-scogo\u002Fskillsh-scraper --skill advisor-orchestrator-worker","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002FSKILL.md",true,[51,57,61,65,69,73],{"path":52,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":55,"sourceHtmlUrl":56},"metadata.json","REFERENCE","参考文档",482,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002Fmetadata.json",{"path":58,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":59,"sourceHtmlUrl":60},"README.md",4341,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002FREADME.md",{"path":62,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":63,"sourceHtmlUrl":64},"references\u002Fadvisor-consult.md",1332,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002Freferences\u002Fadvisor-consult.md",{"path":66,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":67,"sourceHtmlUrl":68},"references\u002Ffallbacks.md",1861,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002Freferences\u002Ffallbacks.md",{"path":70,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":71,"sourceHtmlUrl":72},"references\u002Fworker-brief.md",1198,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002Freferences\u002Fworker-brief.md",{"path":74,"resourceType":53,"resourceTypeDesc":54,"sizeBytes":75,"sourceHtmlUrl":76},"Skill.md",7226,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fshubhamsaboo-awesome-llm-apps\u002Femail-drafter\u002FSkill.md",[78,86,91,98,103,110],{"skillNo":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":82,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":19},"01M7D6MC9FAAEV4CMNYX8W7JTG","arxiv-search","Search arXiv physics, math, and computer science preprints using natural language queries. Powered by Valyu semantic search.","MIT","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fyorkeccak-scientific-skills\u002Farxiv-search\u002Fskills\u002Farxiv-search\u002FSKILL.md",[],[],{"skillNo":87,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":82,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7D6M85CKKN87TA98VCQV4H7","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fyorkeccak-scientific-skills\u002Farxiv-search\u002FSkill.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":82,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M7D4ZTCS4DR3CE4JY06X84T2","working-with-legacy-code","Safely change and test untested codebases using Feathers' \"Working Effectively with Legacy Code\". Use when the user mentions \"legacy code\", \"no tests\", \"untested codebase\", \"how do I test this\", \"seams\", \"characterization tests\", \"golden master\", \"sprout method\", \"afraid to change this code\", \"monster method\", \"dependency breaking\", or \"inherited a messy codebase\". Also trigger when changing code without tests safely, getting a class under test when constructors, statics, or singletons block it, adding features to tangled modules, or planning incremental test coverage for an old codebase. Covers the legacy-code change algorithm, seams, characterization tests, sprout\u002Fwrap, and dependency-breaking techniques. For refactoring code that already has tests, see refactoring-patterns. For day-to-day code quality, see clean-code.","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fworking-with-legacy-code\u002Fworking-with-legacy-code\u002FSKILL.md",[],[],{"skillNo":99,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":82,"licenseFileUrl":6,"sourceHtmlUrl":100,"categories":101,"keywords":102,"publishedAt":19},"01M7D4ZHZKJD2RSNGAF63V14P4","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fworking-with-legacy-code\u002FSkill.md",[],[],{"skillNo":104,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":82,"licenseFileUrl":6,"sourceHtmlUrl":107,"categories":108,"keywords":109,"publishedAt":19},"01M7D4ZB4CJBFW5JE67CPH2QRV","web-typography","Select, pair, and implement typefaces for web projects. Use when the user mentions \"font pairing\", \"which typeface\", \"line height\", \"responsive typography\", \"web font loading\", \"type hierarchy\", \"variable fonts\", \"FOUT\u002FFOIT\", \"typographic scale\", or \"the text is hard to read\". Also trigger when choosing between system fonts and web fonts, optimizing font-loading performance, or designing readable long-form content. Covers readability evaluation, CSS implementation, and performance optimization. For overall UI design systems, see refactoring-ui. For dramatic typographic experiences, see top-design.","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fweb-typography\u002Fweb-typography\u002FSKILL.md",[],[],{"skillNo":111,"name":105,"nameZh":6,"description":106,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":82,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":114,"publishedAt":19},"01M7D4Z4MVYVY3NMJ18BNV6WNA","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fweb-typography\u002FSkill.md",[],[]]