[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M74PFVGND0W8B6EX91RWAVSE":3,"skill-same:01M74PFVGND0W8B6EX91RWAVSE":56},{"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":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":32,"ownerHtmlUrl":33,"ownerBlogUrl":34,"ownerDescription":6,"updatedAt":35,"install":36,"resources":49},"01M74PFVGND0W8B6EX91RWAVSE","compact-docs-writer",null,"Draft, rewrite, or refine a doc for maximum token economy without losing any rule or intent. Use for docs kept in version control and regularly re-read by agents; skip throwaway docs like plans.","eai-org\u002Fagent-toolkit","eai-org","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F301886321?v=4","01M74PFC3HF57T58MQ14P212BQ","Engineering in the age of AI","01M74PFC46MSKTT9SW5Q437FSD",85,"MIT","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Fskills\u002Fcompact-docs-writer\u002FSKILL.md",[],[],"2026-11-11 15:11:14",false,"---\nname: compact-docs-writer\ndescription: Draft, rewrite, or refine a doc for maximum token economy without losing any rule or intent. Use for docs kept in version control and regularly re-read by agents; skip throwaway docs like plans.\nallowed-tools: Read, Write, Edit, Glob, Grep, Bash\nlicense: MIT\nmetadata:\n  version: \"1.8\"\n---\n\n# Compact docs writer\n\nRewrite a doc — or draft a new one — for **token economy**: carry **all** its rules and intent in\nthe **least text possible**, because the doc loads into agent context and is paid for on every\nread. Then prove nothing was lost: present the change with a word delta measured from the files,\nand apply only on approval.\n\n## Core principle\n\nWrite each piece of information with the least text that still preserves every rule, constraint,\nedge case, and intent. Two directions, equally binding:\n\n- Cut duplication, filler, and anything restatable more briefly.\n- **Never** drop text whose removal loses information or instruction, just to be shorter.\n\nRecurring reflex: *\"Can this exact rule be said in fewer words?\"* — if yes, do it.\n\nThe **no-op test** licenses one more deletion. Ask of each sentence, in isolation: *\"does it\nchange the agent's behaviour versus its default?\"* If not, it's a no-op — the agent already acts\nthis way, so removing it loses nothing: delete the whole sentence rather than trimming words from\nit.\n\nCompaction counts words and information density, not whitespace. Blank lines between distinct chunks\ncost effectively nothing and aid the human reader, so keep them where they help; never collapse a\nlong passage into one dense block to look shorter. The same economy runs both ways: a\nhuman-readability gain that is free or near-free in tokens — a blank line, a line break, a\nsemicolon-chained enumeration rendered as a bullet list — is always applied, never skipped to look\ncompact.\n\nStructure follows the same economy: **co-locate** a concept — its rule, exceptions, and caveats\nunder one heading, never scattered — so a reader who jumps to one part gets the others with it.\n\n## Leading words\n\nWhen one concept keeps getting restated, collapse it into a single **leading word** the model\nalready carries from pretraining, and reuse that word wherever the concept applies: it anchors the\nsame behaviour in one token and reads sharper than any paraphrase. The collapse still obeys the\ncore principle — the word must carry every constraint it replaces, and whatever it doesn't carry\nstays spelled out: \"fast, low-overhead feedback\" collapses into a *tight* loop, but a\n\"deterministic\" requirement isn't inside *tight*, so it survives as its own word. Hunt for these\ncollapses in every pass.\n\n## Workflow\n\n1. **Compact.** Rewrite the target to meet the core principle in one pass — a first draft already\n   meets the standard; don't ship a loose draft expecting a later pass to tighten it. If a chunk is\n   needed only in a sub-case and is big enough to tax every read, you *may* suggest extracting it\n   into a referenced file — never force it; the enforced standard is compact text, not splitting.\n2. **Self-review** before presenting (terse yes\u002Fno):\n   - Every original rule, instruction, edge case, and intent still present?\n   - Every surviving rule in the fewest words — tight phrasing, not just free of redundancy?\n     Answer by drafting a shorter rival phrasing for each new or rewritten sentence, not by\n     re-reading: an unchallenged yes is a rubber stamp.\n   - Removal audit against the **rendered diff, not memory**: read every removed line — and every\n     reordered or merged one, which count as removals — and confirm each drops only duplication,\n     filler, or a verified no-op, never a load-bearing rule, instruction, edge case, or nuance.\n     After a merge, re-verify the result still carries every item from both sources.\n   - Addition audit against the same rendered diff: every added sentence must pass the no-op\n     test — cut any that fail.\n   - [scripts\u002Fcheck-doc.sh](scripts\u002Fcheck-doc.sh) passes on the draft, run with\n     `CHECK_DOC_AS=\u003Ctarget path>` while it sits in scratch? Wrap and link conventions a script\n     can hold; fix every failure.\n3. **Present & confirm.** Show the change as a unified diff inside a fenced `diff` code block —\n   every removed line prefixed `-`, every added line `+`, so they render red\u002Fgreen — and when a\n   long line changes by only a few words, add a word-level view (`[-removed-]{+added+}`)\n   pinpointing them. Include a word\u002Ftoken delta **measured from the files, never estimated**:\n   write the not-yet-applied draft to a scratch file (in the session's temp\u002Fscratch dir, never the\n   working tree) and `wc -w` it against the original. Label it not yet applied and awaiting\n   approval; apply only on approval; after applying, say so plainly. Ask for approval in the\n   presentation text or in a later turn, never via a question tool call in the same turn: text\n   emitted before a tool call may not be displayed, so the question would land without the draft.\n","# Compact docs writer\n\nRewrite a doc — or draft a new one — for **token economy**: carry **all** its rules and intent in\nthe **least text possible**, because the doc loads into agent context and is paid for on every\nread. Then prove nothing was lost: present the change with a word delta measured from the files,\nand apply only on approval.\n\n## Core principle\n\nWrite each piece of information with the least text that still preserves every rule, constraint,\nedge case, and intent. Two directions, equally binding:\n\n- Cut duplication, filler, and anything restatable more briefly.\n- **Never** drop text whose removal loses information or instruction, just to be shorter.\n\nRecurring reflex: *\"Can this exact rule be said in fewer words?\"* — if yes, do it.\n\nThe **no-op test** licenses one more deletion. Ask of each sentence, in isolation: *\"does it\nchange the agent's behaviour versus its default?\"* If not, it's a no-op — the agent already acts\nthis way, so removing it loses nothing: delete the whole sentence rather than trimming words from\nit.\n\nCompaction counts words and information density, not whitespace. Blank lines between distinct chunks\ncost effectively nothing and aid the human reader, so keep them where they help; never collapse a\nlong passage into one dense block to look shorter. The same economy runs both ways: a\nhuman-readability gain that is free or near-free in tokens — a blank line, a line break, a\nsemicolon-chained enumeration rendered as a bullet list — is always applied, never skipped to look\ncompact.\n\nStructure follows the same economy: **co-locate** a concept — its rule, exceptions, and caveats\nunder one heading, never scattered — so a reader who jumps to one part gets the others with it.\n\n## Leading words\n\nWhen one concept keeps getting restated, collapse it into a single **leading word** the model\nalready carries from pretraining, and reuse that word wherever the concept applies: it anchors the\nsame behaviour in one token and reads sharper than any paraphrase. The collapse still obeys the\ncore principle — the word must carry every constraint it replaces, and whatever it doesn't carry\nstays spelled out: \"fast, low-overhead feedback\" collapses into a *tight* loop, but a\n\"deterministic\" requirement isn't inside *tight*, so it survives as its own word. Hunt for these\ncollapses in every pass.\n\n## Workflow\n\n1. **Compact.** Rewrite the target to meet the core principle in one pass — a first draft already\n   meets the standard; don't ship a loose draft expecting a later pass to tighten it. If a chunk is\n   needed only in a sub-case and is big enough to tax every read, you *may* suggest extracting it\n   into a referenced file — never force it; the enforced standard is compact text, not splitting.\n2. **Self-review** before presenting (terse yes\u002Fno):\n   - Every original rule, instruction, edge case, and intent still present?\n   - Every surviving rule in the fewest words — tight phrasing, not just free of redundancy?\n     Answer by drafting a shorter rival phrasing for each new or rewritten sentence, not by\n     re-reading: an unchallenged yes is a rubber stamp.\n   - Removal audit against the **rendered diff, not memory**: read every removed line — and every\n     reordered or merged one, which count as removals — and confirm each drops only duplication,\n     filler, or a verified no-op, never a load-bearing rule, instruction, edge case, or nuance.\n     After a merge, re-verify the result still carries every item from both sources.\n   - Addition audit against the same rendered diff: every added sentence must pass the no-op\n     test — cut any that fail.\n   - [scripts\u002Fcheck-doc.sh](scripts\u002Fcheck-doc.sh) passes on the draft, run with\n     `CHECK_DOC_AS=\u003Ctarget path>` while it sits in scratch? Wrap and link conventions a script\n     can hold; fix every failure.\n3. **Present & confirm.** Show the change as a unified diff inside a fenced `diff` code block —\n   every removed line prefixed `-`, every added line `+`, so they render red\u002Fgreen — and when a\n   long line changes by only a few words, add a word-level view (`[-removed-]{+added+}`)\n   pinpointing them. Include a word\u002Ftoken delta **measured from the files, never estimated**:\n   write the not-yet-applied draft to a scratch file (in the session's temp\u002Fscratch dir, never the\n   working tree) and `wc -w` it against the original. Label it not yet applied and awaiting\n   approval; apply only on approval; after applying, say so plainly. Ask for approval in the\n   presentation text or in a later turn, never via a question tool call in the same turn: text\n   emitted before a tool call may not be displayed, so the question would land without the draft.\n",{"name":5,"license":15,"metadata":24,"description":7,"allowed-tools":26},{"version":25},"1.8","Read, Write, Edit, Glob, Grep, Bash","main","12db59437c9d48b9bd62b18888e9974b17adafaf","skills\u002Fcompact-docs-writer",19,"https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit","https:\u002F\u002Feai-org.github.io\u002Fagent-toolkit-docs","https:\u002F\u002Fgithub.com\u002Feai-org","https:\u002F\u002Fmedium.com\u002Fengineering-in-the-age-of-ai","2026-11-14 22:36:21",{"source":8,"skillName":5,"commands":37,"prompt":47,"zipAvailable":48},[38,41,44],{"tool":39,"command":40},"npx","npx skills add eai-org\u002Fagent-toolkit --skill compact-docs-writer",{"tool":42,"command":43},"bunx","bunx skills add eai-org\u002Fagent-toolkit --skill compact-docs-writer",{"tool":45,"command":46},"pnpm","pnpm dlx skills add eai-org\u002Fagent-toolkit --skill compact-docs-writer","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Fskills\u002Fcompact-docs-writer\u002FSKILL.md",true,[50],{"path":51,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":54,"sourceHtmlUrl":55},"scripts\u002Fcheck-doc.sh","SCRIPT","脚本",5633,"https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Fskills\u002Fcompact-docs-writer\u002Fscripts\u002Fcheck-doc.sh",[57,63,70,77,84,91],{"skillNo":58,"name":59,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M74PKAVPTTY5PSSAWHCRVY9T","no-frontmatter","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Ftest\u002Fcompact-skill-creator\u002Ffixtures\u002Fno-frontmatter\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M74PK8J99W7V3669Y4JG2JKV","other-name","Fixture skill whose name differs from its folder.","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Ftest\u002Fcompact-skill-creator\u002Ffixtures\u002Fname-mismatch\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M74PK6T9DHVRCNHTEV2ZA4Y3","md-off","A fixture skill whose folder turns the line length off below a root that sets 99.","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Ftest\u002Fcompact-skill-creator\u002Ffixtures\u002Fmd-off\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M74PK4H4N3D290V4C08M4DGY","long-line","Fixture skill with one prose line of 100 characters, over the fixtures' limit of 99.","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Ftest\u002Fcompact-skill-creator\u002Ffixtures\u002Flong-line\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M74PK2DJ2MCR5KFPMYVQWNJG","flag-without-flow","Fixture skill with the flag and no type.","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Ftest\u002Fcompact-skill-creator\u002Ffixtures\u002Fflag-without-flow\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":15,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M74PK03XGEBQ986FDAE46WCM","double-backtick","A fixture skill with double-backtick code spans holding a link template and an overrun.","https:\u002F\u002Fgithub.com\u002Feai-org\u002Fagent-toolkit\u002Fblob\u002Fmain\u002Ftest\u002Fcompact-skill-creator\u002Ffixtures\u002Fdouble-backtick\u002FSKILL.md",[],[]]