[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79DXGHKCVKJ7WJVEYP7BCCE":3,"skill-same:01M79DXGHKCVKJ7WJVEYP7BCCE":52},{"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":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":29,"ownerDescription":30,"updatedAt":31,"install":32,"resources":45},"01M79DXGHKCVKJ7WJVEYP7BCCE","review-thread",null,"Audit a coding thread, run, or PR against its claims and evidence; return findings and a continuation prompt.","AO-HyS\u002Fdevelopment-system","AO-HyS","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F286554351?v=4","01M79DCXZGC6076EH5QATNXCY2","AO HyS","01M79DCY06TSBCJ946VP0BT0RG",0,"MIT","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F1.34.1\u002Fskills\u002Finternal\u002Freview-thread\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: review-thread\ndescription: Audit a coding thread, run, or PR against its claims and evidence; return findings and a continuation prompt.\n---\n\n# Review a thread\n\nReview completed or ongoing work from outside its implementation context.\nThe selected reviewer can be any available model with the required capabilities;\nthis method has no dependency on a named model or provider. Optimize useful\ncorrection and completion time, not the volume of findings.\n\n## Locate the candidate\n\nUse the supplied thread ID and the host's native history\u002Fexport or local\nread-only history store. Recover the user's outcome, accepted decisions, latest\ncorrections and claimed result. Prefer a bounded snapshot with pointers over\nloading every earlier conversation or handing its full history to a worker.\nOn this workstation, consult [local-history.md](references\u002Flocal-history.md)\nonly when the supplied ID cannot be opened directly in the current host.\n\nVerify the actual root, branch, commit\u002Fdiff and PR. Sidebar links and model\nselection labels can be stale. Associate each finding with the inspected\nrevision. If work changes during the review, record that and inspect the\nrelevant delta before claiming the newest candidate is faulty or accepted.\nAn unsuccessful limited search means “not located”, not “does not exist”.\n\nKeep product files, remote records and active sessions unchanged during audit.\nWrite private review artifacts when useful. A request to review does not itself\nauthorize implementation, external messages, merge or production changes.\n\n## Test the claims\n\nCompare the requested behavior to the diff and the evidence already available.\nPrioritize failures that prevent completion: omitted scope, wrong data\u002Fstate,\nauthorization, broken interactions, visual direction, and unsupported delivery\nclaims. Inspect the exact provider\u002Fruntime contract when correctness hinges on\nit. Tests that mock away a service restriction cannot settle that question.\n\nSeparate confirmed defects, corrected findings and unverified claims. Confirm\ncheck exit codes and what each check covers. Reuse evidence while its revision,\nenvironment and inputs still apply; rerun only checks needed to resolve an\nactual uncertainty. Do not demand another full audit by default.\n\nFor visual claims, open approved references and the actual rendered evidence\nwith a vision-capable reviewer. Check route, role, loaded data, viewport and\ncandidate revision before judging fidelity. Desktop footage is not mobile\nevidence. If browser access or vision is unavailable, report that limit and\ncontinue the review supported by available evidence.\n\nIf runtime verification needs data changes, use an authorized isolated fixture\nand recoverable cleanup. Do not mutate existing user data merely to obtain a\nscreenshot. Keep credentials and auth state out of reports and uploads.\n\n## Inspect orchestration only where it explains the result\n\nUse observed task\u002Fsession records to determine whether workers actually ran,\nwhich models were used, whether independent work overlapped, and where retries\nor repeated investigation occurred. Distinguish a bad parent instruction,\nworker error, environment failure and missing verification. Attribute causes\nonly when the trace supports them; one failed run is not a universal model\nranking. A plan to delegate is not a completed delegation.\n\nMeasure time to the requested usable result. Separate interruptions and provider\nwaits when timestamps support it. Report per-request usage without double\ncounting cumulative snapshots. Distinguish cached tokens, catalog\u002FAPI cost,\nsubscription quota and cash paid; mark unavailable measurements as unknown.\n\n## Return a correction the implementing thread can use\n\nLead with whether the inspected candidate meets the request. For each material\nfinding, give the trigger, observed consequence, exact source\u002Frevision or\nevidence, and the smallest repair acceptance criterion. Explicitly retain\nresolved findings rather than assigning them again. Label missing proof as a\nverification gap rather than inventing a defect.\n\nProvide a short continuation prompt containing the goal, current root\u002Frevision,\nopen findings, useful evidence paths, focused checks and authorized endpoint.\nPreserve working code and settled decisions. Do not broaden it into a redesign\nof the development system, a new benchmark or a mandatory provider change.\nThe verdict and continuation packet complete a review request. Ask for input\nonly when a missing fact prevents a material conclusion; do not end by asking\nwhether to begin an unrequested implementation stage.\n","# Review a thread\n\nReview completed or ongoing work from outside its implementation context.\nThe selected reviewer can be any available model with the required capabilities;\nthis method has no dependency on a named model or provider. Optimize useful\ncorrection and completion time, not the volume of findings.\n\n## Locate the candidate\n\nUse the supplied thread ID and the host's native history\u002Fexport or local\nread-only history store. Recover the user's outcome, accepted decisions, latest\ncorrections and claimed result. Prefer a bounded snapshot with pointers over\nloading every earlier conversation or handing its full history to a worker.\nOn this workstation, consult [local-history.md](references\u002Flocal-history.md)\nonly when the supplied ID cannot be opened directly in the current host.\n\nVerify the actual root, branch, commit\u002Fdiff and PR. Sidebar links and model\nselection labels can be stale. Associate each finding with the inspected\nrevision. If work changes during the review, record that and inspect the\nrelevant delta before claiming the newest candidate is faulty or accepted.\nAn unsuccessful limited search means “not located”, not “does not exist”.\n\nKeep product files, remote records and active sessions unchanged during audit.\nWrite private review artifacts when useful. A request to review does not itself\nauthorize implementation, external messages, merge or production changes.\n\n## Test the claims\n\nCompare the requested behavior to the diff and the evidence already available.\nPrioritize failures that prevent completion: omitted scope, wrong data\u002Fstate,\nauthorization, broken interactions, visual direction, and unsupported delivery\nclaims. Inspect the exact provider\u002Fruntime contract when correctness hinges on\nit. Tests that mock away a service restriction cannot settle that question.\n\nSeparate confirmed defects, corrected findings and unverified claims. Confirm\ncheck exit codes and what each check covers. Reuse evidence while its revision,\nenvironment and inputs still apply; rerun only checks needed to resolve an\nactual uncertainty. Do not demand another full audit by default.\n\nFor visual claims, open approved references and the actual rendered evidence\nwith a vision-capable reviewer. Check route, role, loaded data, viewport and\ncandidate revision before judging fidelity. Desktop footage is not mobile\nevidence. If browser access or vision is unavailable, report that limit and\ncontinue the review supported by available evidence.\n\nIf runtime verification needs data changes, use an authorized isolated fixture\nand recoverable cleanup. Do not mutate existing user data merely to obtain a\nscreenshot. Keep credentials and auth state out of reports and uploads.\n\n## Inspect orchestration only where it explains the result\n\nUse observed task\u002Fsession records to determine whether workers actually ran,\nwhich models were used, whether independent work overlapped, and where retries\nor repeated investigation occurred. Distinguish a bad parent instruction,\nworker error, environment failure and missing verification. Attribute causes\nonly when the trace supports them; one failed run is not a universal model\nranking. A plan to delegate is not a completed delegation.\n\nMeasure time to the requested usable result. Separate interruptions and provider\nwaits when timestamps support it. Report per-request usage without double\ncounting cumulative snapshots. Distinguish cached tokens, catalog\u002FAPI cost,\nsubscription quota and cash paid; mark unavailable measurements as unknown.\n\n## Return a correction the implementing thread can use\n\nLead with whether the inspected candidate meets the request. For each material\nfinding, give the trigger, observed consequence, exact source\u002Frevision or\nevidence, and the smallest repair acceptance criterion. Explicitly retain\nresolved findings rather than assigning them again. Label missing proof as a\nverification gap rather than inventing a defect.\n\nProvide a short continuation prompt containing the goal, current root\u002Frevision,\nopen findings, useful evidence paths, focused checks and authorized endpoint.\nPreserve working code and settled decisions. Do not broaden it into a redesign\nof the development system, a new benchmark or a mandatory provider change.\nThe verdict and continuation packet complete a review request. Ask for input\nonly when a missing fact prevents a material conclusion; do not end by asking\nwhether to begin an unrequested implementation stage.\n",{"name":5,"description":7},"main","9f5b43aadb1a92441d65e9cb74a584cd6fca62bc","artifacts\u002F1.34.1\u002Fskills\u002Finternal\u002Freview-thread","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system","https:\u002F\u002Fgithub.com\u002FAO-HyS","https:\u002F\u002Faohys.net\u002F","Software Boutique, we make your projects come true live. ","2026-11-14 03:17:27",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add AO-HyS\u002Fdevelopment-system --skill review-thread",{"tool":38,"command":39},"bunx","bunx skills add AO-HyS\u002Fdevelopment-system --skill review-thread",{"tool":41,"command":42},"pnpm","pnpm dlx skills add AO-HyS\u002Fdevelopment-system --skill review-thread","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F1.34.1\u002Fskills\u002Finternal\u002Freview-thread\u002FSKILL.md",true,[46],{"path":47,"resourceType":48,"resourceTypeDesc":49,"sizeBytes":50,"sourceHtmlUrl":51},"references\u002Flocal-history.md","REFERENCE","参考文档",1600,"https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F1.34.1\u002Fskills\u002Finternal\u002Freview-thread\u002Freferences\u002Flocal-history.md",[53,60,67,74,81,88],{"skillNo":54,"name":55,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M79EQ4BPYQTX8RWN1Y1J2K1F","working-backwards","Define a software feature through PRFAQ and staged documents when the user requests working backwards before coding.","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.8.0\u002Fskills\u002Finternal\u002Fworking-backwards\u002FSKILL.md",[],[],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":19},"01M79EQ2E8P85Y1Y7VNJZMNEJE","flow-implement","Implement a requested software change through verification, corrections, and its authorized delivery endpoint.","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.8.0\u002Fskills\u002Finternal\u002Fflow-implement\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":19},"01M79EQ0PNMKHPJ78F8WMNX06P","development-steward","Run the scheduled, read-only weekly maintenance review for the six explicitly opted-in repositories and prepare one concise private report for Check-in.","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.8.0\u002Fskills\u002Finternal\u002Fdevelopment-steward\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":19},"01M79EPXJ56MX7955GWP1HD4B0","grill-with-docs","Run one recoverable interview for functional, visual, or mixed software grills; reuse equivalent answers and offer the shared HTML questionnaire.","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.7.0\u002Fskills\u002Finternal\u002Fgrill-with-docs\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":19},"01M79EPTFPHYEV40ZG52Z0GVS2","global-agent-guardrails","Install, audit, test, or update the Development System's global destructive-command guard for Codex, Claude Code and T3 Code. Use when configuring agent safety hooks or investigating whether dangerous shell operations are blocked.","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.7.0\u002Fskills\u002Finternal\u002Fglobal-agent-guardrails\u002FSKILL.md",[],[],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":92,"licenseFileUrl":93,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":19},"01M79EPR0KPHKAFN7V6CEEDR0Y","impeccable","Design or refine frontend UI using project context and a focused design playbook; also handles visual critique and polish.","Apache-2.0","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.7.0\u002Fskills\u002Fexternal\u002Fimpeccable\u002FLICENSE","https:\u002F\u002Fgithub.com\u002FAO-HyS\u002Fdevelopment-system\u002Fblob\u002Fmain\u002Fartifacts\u002F2.7.0\u002Fskills\u002Fexternal\u002Fimpeccable\u002FSKILL.md",[],[]]