[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DATF4KKJS12BBM1VVFW1HF":3,"skill-same:01M6DATF4KKJS12BBM1VVFW1HF":123},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":22,"publishedAt":43,"linkOnly":44,"markdownContent":45,"markdownBody":46,"frontMatter":47,"refName":59,"commitSha":60,"skillDirectory":61,"forksCount":13,"repositoryHtmlUrl":62,"repositoryHomepageUrl":6,"ownerHtmlUrl":63,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":64,"install":65,"resources":78},"01M6DATF4KKJS12BBM1VVFW1HF","th-engineering",null,"Use for engineering-quality work on any change or codebase — holding an implementation or review to the engineering quality bar, code readability, test-suite rigor, CI pipeline discipline (where checks run, duplicate runs, slow or clogged CI), module boundaries and dependency chains, diagnosing hard bugs and flaky failures to root cause, post-refactor cruft cleanup, codebase documentation, scaffolding or reviewing skills against the skill quality bar, greenfield-parity complexity audits that re-run without duplicating beads, or Excalidraw diagrams. Triggers: \"hold this to the engineering bar\", \"is this code readable\", \"review these tests\", \"CI is too slow\", \"why does CI run twice\", \"untangle these dependencies\", \"diagnose this\", \"flaky test\", \"clean up this refactor\", \"would we build it this way from scratch\", \"is this over-engineered\", \"document this service\", \"audit the docs\", \"review this skill\", \"create a skill\", \"draw a diagram\".","tzeusy-org\u002Fai-bootstrap","tzeusy-org","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F141543652?v=4","01M6DAM1AP09H27BYGFY9QCNPB","01M6DAM1F4R187EK11XDN9JZKH",0,"MIT","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-engineering\u002FSKILL.md",[17],{"categoryNo":18,"categoryName":19,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000008","测试与质量","SCENARIO","使用场景",[23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42],"th engineering","engineering","work on","openai codex","scaffolding","clean","diagrams","uv","complexity","up","work","review","beads","pipeline","ci","code","create","cleanup","documentation","change","2026-11-10 10:12:30",false,"---\nname: th-engineering\ndescription: >\n  Use for engineering-quality work on any change or codebase — holding an\n  implementation or review to the engineering quality bar, code readability,\n  test-suite rigor, CI pipeline discipline (where checks run, duplicate\n  runs, slow or clogged CI), module boundaries and dependency chains, diagnosing hard\n  bugs and flaky failures to root cause, post-refactor cruft cleanup,\n  codebase documentation, scaffolding or reviewing skills against the skill\n  quality bar, greenfield-parity complexity audits that re-run without\n  duplicating beads, or Excalidraw diagrams. Triggers: \"hold\n  this to the engineering bar\", \"is this code readable\", \"review these\n  tests\", \"CI is too slow\", \"why does CI run twice\", \"untangle these dependencies\", \"diagnose this\", \"flaky test\",\n  \"clean up this refactor\", \"would we build it this way from scratch\",\n  \"is this over-engineered\", \"document this service\", \"audit the docs\",\n  \"review this skill\", \"create a skill\", \"draw a diagram\".\nmetadata:\n  owner: tze\n  authors:\n    - tze\n    - Claude Fable 5\n    - Claude Fable 5.1\n    - OpenAI Codex\n    - Claude Opus 5.5\n  status: active\n  last_reviewed: \"2026-11-01\"\ncompatibility: skill-standards auditing and excalidraw rendering require uv and Python 3.11+; excalidraw-diagram additionally requires Playwright Chromium setup on first render.\n---\n\n# TH Engineering\n\nSuperskill router for the engineering quality bar. Twelve subskills live under\n`subskills\u002F`, each a complete skill package. **Not** in the global\ncatalog — discover lazily, load **at most one** subskill body per\nsubdomain, and apply several in sequence in your own session when a task\nspans them (\"Subagent dispatch\" says when delegating pays).\n\n`\u002Fth-projects` governs the project (doctrine, specs, topology, audits);\n`\u002Fth-tooling` governs this machine's harness (installed skills, dotfiles,\nsnapshot state); this superskill governs the change: the bar code must clear\nwhile written, reviewed, or cleaned up.\n\n## Discover subskills\n\n```bash\nPKG=\"$(dirname \"\u003Cabsolute-path-to-this-SKILL.md>\")\"\nfind \"$PKG\u002Fsubskills\" -maxdepth 2 -name SKILL.md\nrg -n \"^name:|^description:\" \"$PKG\"\u002Fsubskills\u002F*\u002FSKILL.md\n```\n\n## Routing table\n\n| Task intent | Subskill | Typical trigger |\n|---|---|---|\n| Holistic quality bar: default engineering biases, definition of done; the source projects' craft-and-care pillars adopt by reference. | [subskills\u002Fengineering-bar\u002FSKILL.md](subskills\u002Fengineering-bar\u002FSKILL.md) | \"hold this to the engineering bar\", \"is this change complete\" |\n| Readability and maintainability: naming, function shape, abstraction altitude, comments, simplicity over cleverness. | [subskills\u002Fcode-readability\u002FSKILL.md](subskills\u002Fcode-readability\u002FSKILL.md) | \"is this readable\", \"simplify this code\" |\n| Test quality: behavior-focused assertions, edge\u002Ffailure coverage, regression tests, tautology and flake elimination. | [subskills\u002Ftest-rigor\u002FSKILL.md](subskills\u002Ftest-rigor\u002FSKILL.md) | \"review these tests\", \"what coverage is missing\" |\n| CI\u002FCD pipeline discipline: gate tiers by cost, each verdict once per tree, affected-only PR runs, merge-queue topology, queue-wait vs execution diagnosis, runtime budgets, agent push\u002Frun habits. | [subskills\u002Fci-discipline\u002FSKILL.md](subskills\u002Fci-discipline\u002FSKILL.md) | \"CI is too slow\", \"why does CI run twice\", \"pre-commit or CI?\" |\n| Module boundaries and dependency chains: direction, layering, cycles, public surface, third-party policy. | [subskills\u002Fdependency-hygiene\u002FSKILL.md](subskills\u002Fdependency-hygiene\u002FSKILL.md) | \"untangle these dependencies\", \"should we add this library\" |\n| Codebase documentation: synthesis, diagrams, contract-level code-cited facts, interface semantics cards, human-readable doc sites. | [subskills\u002Fdocumentation\u002FSKILL.md](subskills\u002Fdocumentation\u002FSKILL.md) | \"document this service\", \"audit docs for stale claims\" |\n| Evidence bar for diagnosing hard bugs, flaky failures, and perf regressions: feedback loop, ranked hypotheses, tagged instrumentation, regression test at the correct seam. | [subskills\u002Fdiagnosis\u002FSKILL.md](subskills\u002Fdiagnosis\u002FSKILL.md) | \"diagnose this\", \"fails one run in twenty\", \"find the regression\" |\n| Finish same-repo refactors\u002Frenames\u002Fmigrations: delete lingering aliases, wrappers, fallbacks, dead flags. | [subskills\u002Fcruft-cleanup\u002FSKILL.md](subskills\u002Fcruft-cleanup\u002FSKILL.md) | \"clean up this refactor\", \"old path still works\" |\n| Complexity vs. spec: sketch the from-scratch build, classify every excess as justified or residue, rewrite-vs-incremental; audits that file beads re-run idempotently against the existing backlog and close out with an owner dossier questionnaire (web form URL) for their hard gates. | [subskills\u002Fgreenfield-parity\u002FSKILL.md](subskills\u002Fgreenfield-parity\u002FSKILL.md) | \"would we build it this way from scratch\", \"is this over-engineered\", \"re-run the audit without duplicating beads\" |\n| Skill\u002Fsuperskill quality bar: triggers, grounding, metadata, routing, context efficiency, validation. | [subskills\u002Fskill-standards\u002FSKILL.md](subskills\u002Fskill-standards\u002FSKILL.md) | \"review this skill\", \"should this be a superskill\" |\n| Scaffold or author a new skill end-to-end: init from template, bundle scripts\u002Freferences\u002Fassets, package for distribution. | [subskills\u002Fskill-creator\u002FSKILL.md](subskills\u002Fskill-creator\u002FSKILL.md) | \"create a new skill\", \"scaffold a skill\", \"package this skill\" |\n| Excalidraw diagrams into\u002Fout of workflows, architectures, protocols, Mermaid. | [subskills\u002Fexcalidraw-diagram\u002FSKILL.md](subskills\u002Fexcalidraw-diagram\u002FSKILL.md) | \"draw a diagram\", \"convert this Mermaid\" |\n\n## Routing rules\n\n- **Bar vs. subdomain**: holistic \"is this good\u002Fdone\" → engineering-bar; an\n  ask naming one subdomain → that subskill directly.\n- **Test worth vs. pipeline placement**: whether a test should exist, suite\n  tiers, condensation → test-rigor; where and how often checks run, slow or\n  duplicated CI, merge queues → ci-discipline (which sends condensation back\n  to test-rigor as its last rung). A red or flaky run → diagnosis.\n- **Change-level vs. project-level**: changes, diffs, PRs, modules → here.\n  Repo-wide audits, specs, prioritization, knowledge architecture →\n  `\u002Fth-projects` (whose craft-and-care pillar adopts engineering-bar by\n  reference, and whose project-review cites greenfield-parity as its\n  complexity lens).\n- **Refactor leftovers vs. architecture residue**: wrappers and fallbacks\n  from one finished migration → cruft-cleanup; whole surfaces the current\n  spec would never have produced → greenfield-parity.\n- **Build quality vs. product feel**: how the code is built → here; how the\n  product feels to use (UX contracts, perceived latency, copy,\n  discoverability, accessibility) → `\u002Fth-design`. A review that surfaces a\n  user-facing feel finding routes it there instead of absorbing it.\n- **Docs craft vs. knowledge architecture**: README\u002Fdocs\u002Fdoc-site quality →\n  documentation (consumes excalidraw-diagram for its diagrams — an\n  expected pairing); five-pillar shape work → `\u002Fth-projects`.\n- **Skill content vs. harness state**: authoring a skill from scratch →\n  skill-creator (then skill-standards before shipping); reviewing or\n  auditing an existing one → skill-standards; whether installed skills are used, linked, fresh on\n  this machine → `\u002Fth-tooling` (audit-skill-hygiene, refresh-snapshots).\n- **Fallback**: quality-adjacent but no row fits → answer from router-level\n  context or ask; don't load a subskill to browse.\n\n## Subagent dispatch\n\nDefault inline. A sweep across several subdomains reads the same diff, so\none warm session applying each subskill in turn beats N cold subagents that\neach re-read it at full price (cache-first rules:\n[`beads-orchestration\u002Freferences\u002Ftoken-efficiency.md`](..\u002Fbeads-orchestration\u002Freferences\u002Ftoken-efficiency.md)).\nDelegate only when one of these holds:\n\n- **Independence**: the reviewer must not be the implementer. A\n  fresh-context verifier catches what self-critique misses on high-risk\n  changes.\n- **Context isolation**: the scope would flood your context with material\n  the rest of the task never needs (repo-wide sweep, large generated\n  surface).\n- **Iteration-heavy single subdomain** (excalidraw render loop, skill\n  audit-and-fix): delegate the whole loop, review the returned artifact.\n\nAt most 3 subagent sessions in flight; meanwhile keep working on anything\nthat does not move the head under review. Each review prompt carries (1) the\nabsolute path to its `subskills\u002F\u003Cname>\u002FSKILL.md` with instruction to read\nand apply it, (2) exact scope (diff, files, directories), (3) what the\nresult feeds and what done looks like, (4) the output contract: every\nevidenced finding with file:line, severity, confidence, and a proposed fix,\nor a clean verdict citing what was checked. The parent filters, dedupes,\nand synthesizes; conflicts resolve via engineering-bar's biases. A loop\ndelegation states the artifact and its done-condition instead of (4).\n\n## Shared invariants (all subskills)\n\n- Quality claims are reviewable expectations, not taste: every finding cites\n  file:line (or skill-path) evidence and the expectation it violates.\n- engineering-bar's default biases are the baseline all subskills assume; a\n  project's `about\u002Fcraft-and-care\u002F` overrides them where present.\n- Fix-it-now beats file-it-away: the current implementation\u002Frecovery owner fixes\n  small in-scope findings. Independent reviewers retain verifier ownership;\n  new architecture, trust-boundary, contract, or risk-class findings return to\n  `\u002Fth-projects` allocation\u002Fspec gates instead of silently expanding the diff.\n- The owner's instructions outrank skill text: a subskill's bar shapes how\n  work is done, not whether a direct request from the human owner (not a\n  dispatching agent) is honored. Safety hard stops stay. When a gate here\n  makes you pause, name the rule and its file.\n- Reports and subagent briefs are written for a cold reader: outcome first,\n  complete sentences, terms defined. The terse prose of these skill files\n  is a loading-cost choice, not a template for deliverables.\n- Subskills reference each other by relative path (`..\u002Fengineering-bar\u002F…`);\n  those paths are package-internal and stable.\n","# TH Engineering\n\nSuperskill router for the engineering quality bar. Twelve subskills live under\n`subskills\u002F`, each a complete skill package. **Not** in the global\ncatalog — discover lazily, load **at most one** subskill body per\nsubdomain, and apply several in sequence in your own session when a task\nspans them (\"Subagent dispatch\" says when delegating pays).\n\n`\u002Fth-projects` governs the project (doctrine, specs, topology, audits);\n`\u002Fth-tooling` governs this machine's harness (installed skills, dotfiles,\nsnapshot state); this superskill governs the change: the bar code must clear\nwhile written, reviewed, or cleaned up.\n\n## Discover subskills\n\n```bash\nPKG=\"$(dirname \"\u003Cabsolute-path-to-this-SKILL.md>\")\"\nfind \"$PKG\u002Fsubskills\" -maxdepth 2 -name SKILL.md\nrg -n \"^name:|^description:\" \"$PKG\"\u002Fsubskills\u002F*\u002FSKILL.md\n```\n\n## Routing table\n\n| Task intent | Subskill | Typical trigger |\n|---|---|---|\n| Holistic quality bar: default engineering biases, definition of done; the source projects' craft-and-care pillars adopt by reference. | [subskills\u002Fengineering-bar\u002FSKILL.md](subskills\u002Fengineering-bar\u002FSKILL.md) | \"hold this to the engineering bar\", \"is this change complete\" |\n| Readability and maintainability: naming, function shape, abstraction altitude, comments, simplicity over cleverness. | [subskills\u002Fcode-readability\u002FSKILL.md](subskills\u002Fcode-readability\u002FSKILL.md) | \"is this readable\", \"simplify this code\" |\n| Test quality: behavior-focused assertions, edge\u002Ffailure coverage, regression tests, tautology and flake elimination. | [subskills\u002Ftest-rigor\u002FSKILL.md](subskills\u002Ftest-rigor\u002FSKILL.md) | \"review these tests\", \"what coverage is missing\" |\n| CI\u002FCD pipeline discipline: gate tiers by cost, each verdict once per tree, affected-only PR runs, merge-queue topology, queue-wait vs execution diagnosis, runtime budgets, agent push\u002Frun habits. | [subskills\u002Fci-discipline\u002FSKILL.md](subskills\u002Fci-discipline\u002FSKILL.md) | \"CI is too slow\", \"why does CI run twice\", \"pre-commit or CI?\" |\n| Module boundaries and dependency chains: direction, layering, cycles, public surface, third-party policy. | [subskills\u002Fdependency-hygiene\u002FSKILL.md](subskills\u002Fdependency-hygiene\u002FSKILL.md) | \"untangle these dependencies\", \"should we add this library\" |\n| Codebase documentation: synthesis, diagrams, contract-level code-cited facts, interface semantics cards, human-readable doc sites. | [subskills\u002Fdocumentation\u002FSKILL.md](subskills\u002Fdocumentation\u002FSKILL.md) | \"document this service\", \"audit docs for stale claims\" |\n| Evidence bar for diagnosing hard bugs, flaky failures, and perf regressions: feedback loop, ranked hypotheses, tagged instrumentation, regression test at the correct seam. | [subskills\u002Fdiagnosis\u002FSKILL.md](subskills\u002Fdiagnosis\u002FSKILL.md) | \"diagnose this\", \"fails one run in twenty\", \"find the regression\" |\n| Finish same-repo refactors\u002Frenames\u002Fmigrations: delete lingering aliases, wrappers, fallbacks, dead flags. | [subskills\u002Fcruft-cleanup\u002FSKILL.md](subskills\u002Fcruft-cleanup\u002FSKILL.md) | \"clean up this refactor\", \"old path still works\" |\n| Complexity vs. spec: sketch the from-scratch build, classify every excess as justified or residue, rewrite-vs-incremental; audits that file beads re-run idempotently against the existing backlog and close out with an owner dossier questionnaire (web form URL) for their hard gates. | [subskills\u002Fgreenfield-parity\u002FSKILL.md](subskills\u002Fgreenfield-parity\u002FSKILL.md) | \"would we build it this way from scratch\", \"is this over-engineered\", \"re-run the audit without duplicating beads\" |\n| Skill\u002Fsuperskill quality bar: triggers, grounding, metadata, routing, context efficiency, validation. | [subskills\u002Fskill-standards\u002FSKILL.md](subskills\u002Fskill-standards\u002FSKILL.md) | \"review this skill\", \"should this be a superskill\" |\n| Scaffold or author a new skill end-to-end: init from template, bundle scripts\u002Freferences\u002Fassets, package for distribution. | [subskills\u002Fskill-creator\u002FSKILL.md](subskills\u002Fskill-creator\u002FSKILL.md) | \"create a new skill\", \"scaffold a skill\", \"package this skill\" |\n| Excalidraw diagrams into\u002Fout of workflows, architectures, protocols, Mermaid. | [subskills\u002Fexcalidraw-diagram\u002FSKILL.md](subskills\u002Fexcalidraw-diagram\u002FSKILL.md) | \"draw a diagram\", \"convert this Mermaid\" |\n\n## Routing rules\n\n- **Bar vs. subdomain**: holistic \"is this good\u002Fdone\" → engineering-bar; an\n  ask naming one subdomain → that subskill directly.\n- **Test worth vs. pipeline placement**: whether a test should exist, suite\n  tiers, condensation → test-rigor; where and how often checks run, slow or\n  duplicated CI, merge queues → ci-discipline (which sends condensation back\n  to test-rigor as its last rung). A red or flaky run → diagnosis.\n- **Change-level vs. project-level**: changes, diffs, PRs, modules → here.\n  Repo-wide audits, specs, prioritization, knowledge architecture →\n  `\u002Fth-projects` (whose craft-and-care pillar adopts engineering-bar by\n  reference, and whose project-review cites greenfield-parity as its\n  complexity lens).\n- **Refactor leftovers vs. architecture residue**: wrappers and fallbacks\n  from one finished migration → cruft-cleanup; whole surfaces the current\n  spec would never have produced → greenfield-parity.\n- **Build quality vs. product feel**: how the code is built → here; how the\n  product feels to use (UX contracts, perceived latency, copy,\n  discoverability, accessibility) → `\u002Fth-design`. A review that surfaces a\n  user-facing feel finding routes it there instead of absorbing it.\n- **Docs craft vs. knowledge architecture**: README\u002Fdocs\u002Fdoc-site quality →\n  documentation (consumes excalidraw-diagram for its diagrams — an\n  expected pairing); five-pillar shape work → `\u002Fth-projects`.\n- **Skill content vs. harness state**: authoring a skill from scratch →\n  skill-creator (then skill-standards before shipping); reviewing or\n  auditing an existing one → skill-standards; whether installed skills are used, linked, fresh on\n  this machine → `\u002Fth-tooling` (audit-skill-hygiene, refresh-snapshots).\n- **Fallback**: quality-adjacent but no row fits → answer from router-level\n  context or ask; don't load a subskill to browse.\n\n## Subagent dispatch\n\nDefault inline. A sweep across several subdomains reads the same diff, so\none warm session applying each subskill in turn beats N cold subagents that\neach re-read it at full price (cache-first rules:\n[`beads-orchestration\u002Freferences\u002Ftoken-efficiency.md`](..\u002Fbeads-orchestration\u002Freferences\u002Ftoken-efficiency.md)).\nDelegate only when one of these holds:\n\n- **Independence**: the reviewer must not be the implementer. A\n  fresh-context verifier catches what self-critique misses on high-risk\n  changes.\n- **Context isolation**: the scope would flood your context with material\n  the rest of the task never needs (repo-wide sweep, large generated\n  surface).\n- **Iteration-heavy single subdomain** (excalidraw render loop, skill\n  audit-and-fix): delegate the whole loop, review the returned artifact.\n\nAt most 3 subagent sessions in flight; meanwhile keep working on anything\nthat does not move the head under review. Each review prompt carries (1) the\nabsolute path to its `subskills\u002F\u003Cname>\u002FSKILL.md` with instruction to read\nand apply it, (2) exact scope (diff, files, directories), (3) what the\nresult feeds and what done looks like, (4) the output contract: every\nevidenced finding with file:line, severity, confidence, and a proposed fix,\nor a clean verdict citing what was checked. The parent filters, dedupes,\nand synthesizes; conflicts resolve via engineering-bar's biases. A loop\ndelegation states the artifact and its done-condition instead of (4).\n\n## Shared invariants (all subskills)\n\n- Quality claims are reviewable expectations, not taste: every finding cites\n  file:line (or skill-path) evidence and the expectation it violates.\n- engineering-bar's default biases are the baseline all subskills assume; a\n  project's `about\u002Fcraft-and-care\u002F` overrides them where present.\n- Fix-it-now beats file-it-away: the current implementation\u002Frecovery owner fixes\n  small in-scope findings. Independent reviewers retain verifier ownership;\n  new architecture, trust-boundary, contract, or risk-class findings return to\n  `\u002Fth-projects` allocation\u002Fspec gates instead of silently expanding the diff.\n- The owner's instructions outrank skill text: a subskill's bar shapes how\n  work is done, not whether a direct request from the human owner (not a\n  dispatching agent) is honored. Safety hard stops stay. When a gate here\n  makes you pause, name the rule and its file.\n- Reports and subagent briefs are written for a cold reader: outcome first,\n  complete sentences, terms defined. The terse prose of these skill files\n  is a loading-cost choice, not a template for deliverables.\n- Subskills reference each other by relative path (`..\u002Fengineering-bar\u002F…`);\n  those paths are package-internal and stable.\n",{"name":5,"metadata":48,"description":57,"compatibility":58},{"owner":49,"status":50,"authors":51,"last_reviewed":56},"tze","active",[49,52,53,54,55],"Claude Fable 5","Claude Fable 5.1","OpenAI Codex","Claude Opus 5.5","2026-11-01","Use for engineering-quality work on any change or codebase — holding an implementation or review to the engineering quality bar, code readability, test-suite rigor, CI pipeline discipline (where checks run, duplicate runs, slow or clogged CI), module boundaries and dependency chains, diagnosing hard bugs and flaky failures to root cause, post-refactor cruft cleanup, codebase documentation, scaffolding or reviewing skills against the skill quality bar, greenfield-parity complexity audits that re-run without duplicating beads, or Excalidraw diagrams. Triggers: \"hold this to the engineering bar\", \"is this code readable\", \"review these tests\", \"CI is too slow\", \"why does CI run twice\", \"untangle these dependencies\", \"diagnose this\", \"flaky test\", \"clean up this refactor\", \"would we build it this way from scratch\", \"is this over-engineered\", \"document this service\", \"audit the docs\", \"review this skill\", \"create a skill\", \"draw a diagram\".\n","skill-standards auditing and excalidraw rendering require uv and Python 3.11+; excalidraw-diagram additionally requires Playwright Chromium setup on first render.","main","c907d6f0bf96bd61b936e8972e0f1c66ebb8148c","skills\u002Fpersonal\u002Fth-engineering","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap","https:\u002F\u002Fgithub.com\u002Ftzeusy-org","2026-11-10 23:17:25",{"source":8,"skillName":5,"commands":66,"prompt":76,"zipAvailable":77},[67,70,73],{"tool":68,"command":69},"npx","npx skills add tzeusy-org\u002Fai-bootstrap --skill th-engineering",{"tool":71,"command":72},"bunx","bunx skills add tzeusy-org\u002Fai-bootstrap --skill th-engineering",{"tool":74,"command":75},"pnpm","pnpm dlx skills add tzeusy-org\u002Fai-bootstrap --skill th-engineering","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-engineering\u002FSKILL.md",true,[79,85,91,95,99,103,107,111,115,119],{"path":80,"resourceType":81,"resourceTypeDesc":82,"sizeBytes":83,"sourceHtmlUrl":84},"condition-based-waiting-example.ts","SCRIPT","脚本",5054,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Fcondition-based-waiting-example.ts",{"path":86,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":89,"sourceHtmlUrl":90},"condition-based-waiting.md","REFERENCE","参考文档",3516,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Fcondition-based-waiting.md",{"path":92,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":93,"sourceHtmlUrl":94},"CREATION-LOG.md",4268,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002FCREATION-LOG.md",{"path":96,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":97,"sourceHtmlUrl":98},"defense-in-depth.md",3650,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Fdefense-in-depth.md",{"path":100,"resourceType":81,"resourceTypeDesc":82,"sizeBytes":101,"sourceHtmlUrl":102},"find-polluter.sh",1528,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Ffind-polluter.sh",{"path":104,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":105,"sourceHtmlUrl":106},"root-cause-tracing.md",5327,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Froot-cause-tracing.md",{"path":108,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":109,"sourceHtmlUrl":110},"test-academic.md",653,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Ftest-academic.md",{"path":112,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":113,"sourceHtmlUrl":114},"test-pressure-1.md",1900,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Ftest-pressure-1.md",{"path":116,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":117,"sourceHtmlUrl":118},"test-pressure-2.md",2283,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Ftest-pressure-2.md",{"path":120,"resourceType":87,"resourceTypeDesc":88,"sizeBytes":121,"sourceHtmlUrl":122},"test-pressure-3.md",2692,"https:\u002F\u002Fgithub.com\u002Fsapsapshen\u002Fai-memory-router\u002Fblob\u002Fmain\u002Fskills-store\u002Fglobal\u002Fsystematic-debugging\u002Ftest-pressure-3.md",[124,150,172,187,205,216],{"skillNo":125,"name":126,"nameZh":6,"description":127,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":128,"categories":129,"keywords":130,"publishedAt":43},"01M6DAZD21TWH2ET70V9FF75CQ","user-run","Use when a step needs privileges this session lacks (docker, sudo, another user's checkout or secrets, host access) and the owner must run it as tze. Writes a reviewable \u002Fhome\u002Forca\u002F.tmp\u002F\u003Cslug>\u002Frun.sh that impersonates tze (USER_TO_IMPERSONATE overrides), tees timestamped logs plus latest.log and collects artifacts there; Claude presents it with a risk and time brief, then continues from the logs. Triggers: \"I'll run it for you\", \"permission denied on docker.sock\", \"needs sudo\", \"sudo: a password is required\", \"run as tze\", \"\u002Fuser-run\".","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fuser-run\u002FSKILL.md",[],[131,132,133,134,135,136,137,138,139,140,50,141,142,143,144,145,146,147,148,149],"user run","user","run","docker","logs","use","host","as","triggers","permission","time","log","artifacts","session","password","access","secrets","risk","brief",{"skillNo":151,"name":152,"nameZh":6,"description":153,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":154,"categories":155,"keywords":164,"publishedAt":43},"01M6DAZAHTCBAD2WF0GAHWPST8","requesting-code-review","Use when completing tasks, implementing major features, or before merging to verify work meets requirements","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-workflow\u002Fsubskills\u002Frequesting-code-review\u002FSKILL.md",[156,159],{"categoryNo":157,"categoryName":158,"taxonomyType":20,"taxonomyTypeDesc":21},"01JBSEEDCAT000000000000007","代码审查",{"categoryNo":160,"categoryName":161,"taxonomyType":162,"taxonomyTypeDesc":163},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[165,38,34,166,33,167,50,168,169,136,170,171],"requesting code review","code review","requirements","upstream","superpowers","verify","tasks",{"skillNo":173,"name":174,"nameZh":6,"description":175,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":176,"categories":177,"keywords":180,"publishedAt":43},"01M6DAZ7CJ82D81XZB2YHVVY48","receiving-code-review","Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-workflow\u002Fsubskills\u002Freceiving-code-review\u002FSKILL.md",[178,179],{"categoryNo":157,"categoryName":158,"taxonomyType":20,"taxonomyTypeDesc":21},{"categoryNo":160,"categoryName":161,"taxonomyType":162,"taxonomyTypeDesc":163},[181,38,34,166,50,168,182,169,136,183,184,185,186],"receiving code review","implementation","code review feedback","technical","feedback","verification",{"skillNo":188,"name":189,"nameZh":6,"description":190,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":191,"categories":192,"keywords":193,"publishedAt":43},"01M6DAZ4ER76ADXJ9KRFP7V3Q9","finishing-a-development-branch","Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-workflow\u002Fsubskills\u002Ffinishing-a-development-branch\u002FSKILL.md",[],[194,195,196,197,136,198,33,199,40,200,169,201,202,203,50,168,204,182],"branch","development","finishing a development branch","integrate","decide","how","pr","merge","tests","structured","options",{"skillNo":206,"name":207,"nameZh":6,"description":208,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":209,"categories":210,"keywords":211,"publishedAt":43},"01M6DAYZFWWW2YPKP89QNQR96H","executing-plans","Use when you have a written implementation plan to execute in a separate session with review checkpoints","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-workflow\u002Fsubskills\u002Fexecuting-plans\u002FSKILL.md",[],[212,213,214,34,215,50,168,182,169,136,144],"plans","executing plans","plan","execute",{"skillNo":217,"name":218,"nameZh":6,"description":219,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":220,"categories":221,"keywords":222,"publishedAt":43},"01M6DAYWM2YQHYCK7C35K1D29J","dispatching-parallel-agents","Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies","https:\u002F\u002Fgithub.com\u002Ftzeusy-org\u002Fai-bootstrap\u002Fblob\u002Fmain\u002Fskills\u002Fpersonal\u002Fth-workflow\u002Fsubskills\u002Fdispatching-parallel-agents\u002FSKILL.md",[],[223,224,225,136,169,168,226,50,227,171,228],"parallel agents","parallel","dispatching parallel agents","state","dependencies","shared"]