[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EKF6SRPB9FD3R6Q38DBDFE":3,"skill-same:01M7EKF6SRPB9FD3R6Q38DBDFE":72},{"skillNo":4,"name":5,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":20,"frontMatter":21,"refName":22,"commitSha":23,"skillDirectory":24,"forksCount":13,"repositoryHtmlUrl":25,"repositoryHomepageUrl":6,"ownerHtmlUrl":26,"ownerBlogUrl":27,"ownerDescription":28,"updatedAt":29,"install":30,"resources":43},"01M7EKF6SRPB9FD3R6Q38DBDFE","lean-progressive-context",null,"raillen\u002Fdaisy-code-editor","raillen","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F50306048?v=4","01M7EKBDVVBSPF21PBPVJYMNYR","Raillen Santos","01M7EKBDWD12A1K9N9Y638FGX9",0,"MIT","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"# Lean Progressive Context\n\n## Purpose\nSelect the smallest sufficient context for a task under strict token budgets through staged loading: start from a minimal working set, expand one layer at a time with a running token ledger, and stop at the first layer that answers the task.\n\n## Use when\n- Starting any Goal or Task where the relevant codebase exceeds what fits comfortably in the context window.\n- Deciding what to load next (source files, specs, ADRs, test fixtures) instead of dumping whole directories.\n- Defining or enforcing token budgets, expansion caps, and stopping conditions for agent runs.\n- Writing Working Context Capsules that summarize completed stages for downstream agents.\n\n## Do not use when\n- Compressing or rewriting context that is already loaded (use `context-optimization`).\n- Authoring prompts or eval harnesses (use `prompt-engineering`).\n- Performing the domain work itself once context is sufficient; switch to the domain skill.\n\n## Required context\n- Task statement with explicit acceptance criteria.\n- Total token budget for the run (default 60,000 tokens) and per-stage expansion cap (default 8,000 tokens).\n- Entry-point hints: file paths, symbols, or specs most likely relevant.\n- Stopping condition: what evidence proves the current layer is sufficient.\n\n## Procedure\n1. **Declare the budget up front**: record total budget (e.g. 60,000 tokens), per-stage cap (8,000 tokens), and max expansion stages (5) in the run ledger before loading anything.\n2. **Load the minimal working set first**: task statement, entry-point file, and directly referenced spec only; measure with a tokenizer and log the running total.\n3. **Expand one layer at a time**: each expansion names a hypothesis (e.g. \"auth failure originates in token refresh, load auth\u002Frefresh.go next\"), loads only that layer, and appends cost to the ledger.\n4. **Test sufficiency after every layer**: ask whether acceptance criteria can now be addressed; if yes, stop expanding and start domain work.\n5. **Write a capsule on stage completion**: a 10-line Working Context Capsule (goal, decisions, open questions, next layer if needed) so downstream agents inherit understanding without reloading raw context.\n6. **Run `scripts\u002Fverify.sh`** to confirm ledger discipline and capsule presence before handoff.\n\n## Decision rules\n- **Budget first, loading second**: no file is read before its cost is chargeable against a declared budget.\n- **One hypothesis per expansion**: each layer must state what it expects to find; fishing expeditions without hypotheses are prohibited.\n- **Stop at sufficiency**: expanding past the layer that answers the task is a violation, not thoroughness.\n- **Capsules over raw dumps**: handoffs carry capsules, never unbounded file lists.\n- **Ledger honesty**: every loaded file and its token cost appears in the ledger; unlogged loads are treated as budget overruns.\n\n## Evidence required\n- Token ledger listing every loaded file with per-file and running-total costs.\n- Working Context Capsule for each completed stage.\n- Sufficiency note stating which layer answered the task and why expansion stopped.\n- Passing execution log from `scripts\u002Fverify.sh`.\n\n## Output contract\n- Minimal sufficient working set actually used for the domain work.\n- Token ledger proving spend stayed within budget and caps.\n- Capsule chain enabling any reviewer to reconstruct the reasoning path.\n\n## Stop conditions\n- Acceptance criteria addressable with the current layer; expansion halts.\n- Max expansion stages (default 5) reached without sufficiency.\n- Token budget exhausted.\n- Blocked on external dependency.\n\n## Escalation rules\n- Escalate to the task owner if sufficiency is unreachable within budget and scope reduction is needed.\n- Escalate to a human if the entry-point hints are missing and blind search would burn the budget.\n",{},"main","2e465213b972294984c8c9515a47f811e4b358da",".agents\u002Fskills\u002Flean-progressive-context","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor","https:\u002F\u002Fgithub.com\u002Fraillen","https:\u002F\u002Fraillen.com","\u003C\u003CMarketing & IT>>\r\n\r\n@poppy-team\r\n@wasd-lat ","2026-10-16 09:17:52",{"source":7,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add raillen\u002Fdaisy-code-editor --skill lean-progressive-context",{"tool":36,"command":37},"bunx","bunx skills add raillen\u002Fdaisy-code-editor --skill lean-progressive-context",{"tool":39,"command":40},"pnpm","pnpm dlx skills add raillen\u002Fdaisy-code-editor --skill lean-progressive-context","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002FSKILL.md",true,[44,50,54,58,64,68],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"checks\u002Flean-progressive-context-checklist.md","REFERENCE","参考文档",1573,"https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002Fchecks\u002Flean-progressive-context-checklist.md",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"manifest.json",2064,"https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002Fmanifest.json",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Flean-progressive-context-guide.md",2524,"https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002Freferences\u002Flean-progressive-context-guide.md",{"path":59,"resourceType":60,"resourceTypeDesc":61,"sizeBytes":62,"sourceHtmlUrl":63},"scripts\u002Fverify.sh","SCRIPT","脚本",4026,"https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002Fscripts\u002Fverify.sh",{"path":65,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":66,"sourceHtmlUrl":67},"templates\u002Flean-progressive-context-spec.md",1670,"https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002Ftemplates\u002Flean-progressive-context-spec.md",{"path":69,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":70,"sourceHtmlUrl":71},"templates\u002Flean-progressive-context-template.md",826,"https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Flean-progressive-context\u002Ftemplates\u002Flean-progressive-context-template.md",[73,79,85,92,99,106],{"skillNo":74,"name":75,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":76,"categories":77,"keywords":78,"publishedAt":18},"01M7EKQYSDAX3SGSCRF873Z8E2","zoom-reflow","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.ai\u002Fskills\u002Fzoom-reflow\u002FSKILL.md",[],[],{"skillNo":80,"name":81,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":18},"01M7EKQX0MWJNQD482GGZPGES4","wireframe-styleguide","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.ai\u002Fskills\u002Fwireframe-styleguide\u002FSKILL.md",[],[],{"skillNo":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":18},"01M7EKQS57FS6JYE6KWW6X5ADJ","visual-regression","Pixel-by-pixel snapshot testing, Playwright integration, dynamic content masking, deterministic font rendering, and cross-browser CI gates","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.ai\u002Fskills\u002Fvisual-regression\u002FSKILL.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":18},"01M7EKQMFR7651NVH100P5781Z","ui-implementation","Component-driven UI implementation, multi-state matrix (default, hover, focus-visible, active, disabled, loading, empty, error), DTCG design tokens, responsive layout contracts, and accessibility tree mapping.","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.ai\u002Fskills\u002Fui-implementation\u002FSKILL.md",[],[],{"skillNo":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":18},"01M7EKQG3NVX6DPCYS6Q4PA71V","tui-editor-engine","Text buffer and editing semantics for a terminal code editor — index conversion, transaction-first editing, multi-cursor invariants, undo grouping, and large-file behaviour.","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.ai\u002Fskills\u002Ftui-editor-engine\u002FSKILL.md",[],[],{"skillNo":107,"name":108,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":7,"ownerLogin":8,"ownerAvatarUrl":9,"ownerAccountNo":10,"ownerDisplayName":11,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":18},"01M7EKQEJ993JGC72FE2KFY3ZZ","tree-sitter-context-indexing","https:\u002F\u002Fgithub.com\u002Fraillen\u002Fdaisy-code-editor\u002Fblob\u002Fmain\u002F.ai\u002Fskills\u002Ftree-sitter-context-indexing\u002FSKILL.md",[],[]]