[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6DDS85A3RAEZRTYE3ZXS88E":3,"skill-same:01M6DDS85A3RAEZRTYE3ZXS88E":145},{"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":23,"publishedAt":44,"linkOnly":45,"markdownContent":46,"markdownBody":47,"frontMatter":48,"refName":49,"commitSha":50,"skillDirectory":51,"forksCount":14,"repositoryHtmlUrl":52,"repositoryHomepageUrl":6,"ownerHtmlUrl":53,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":54,"install":55,"resources":68},"01M6DDS85A3RAEZRTYE3ZXS88E","acceptance-review",null,"Use at the \"done\" boundary of an app change, before a PR or a \"fixed \u002F it works now\" claim. Dispatches an INDEPENDENT cross-model reviewer (Codex\u002FGemini CLI, else fresh Claude) that adversarially judges whether the real end-to-end journey works or is green-but-broken, over the diff + golden-journeys + on-disk evidence. Returns a SHIP\u002FBLOCK verdict. NOT for code review (use code-review).","Prekzursil\u002Fagent-skills-toolchain","Prekzursil","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54636077?v=4","01M6DACZ1PCQHD99SCR1YXG28B","Andrei Mihai Visalon","01M6DACZ670WZF535V3WYYSS4Z",0,"MIT","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Facceptance-review\u002FSKILL.md",[18],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000007","代码审查","SCENARIO","使用场景",[24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43],"review","acceptance","acceptance review","app","code review","end","green","done","claim","journey","evidence","diff","use","broken","model","disk","code-review","code","pr","gemini","2026-11-10 10:12:30",false,"---\nname: acceptance-review\ndescription: Use at the \"done\" boundary of an app change, before a PR or a \"fixed \u002F it works now\" claim. Dispatches an INDEPENDENT cross-model reviewer (Codex\u002FGemini CLI, else fresh Claude) that adversarially judges whether the real end-to-end journey works or is green-but-broken, over the diff + golden-journeys + on-disk evidence. Returns a SHIP\u002FBLOCK verdict. NOT for code review (use code-review).\n---\n\n# Acceptance Review — the cross-model \"does it actually work?\" verdict\n\nThe external-signal verifier for the \"done\" boundary. A capable model optimizes its\nown done-signal — coverage, \"tests pass\" — and stops, while the real product is\nbroken (dead buttons, \"shorts don't generate\", a severed A→B→C flow). The signal is\ngameable because the *same context that wrote the code certifies it*: intrinsic\nself-correction degrades (Huang, ICLR'24), verifier↔generator errors couple\n(\"Verification Mirage\"), and judges prefer their own output (Panickssery, NeurIPS'24).\nThis skill breaks that loop with **verifier separation** — an evaluator that did NOT\nwrite the code and, when possible, is a **different model** (VeriMAP \u002F MetaGPT), which\ndemands the real artifact and is forbidden from trusting coverage.\n\nIt complements the always-on functional-acceptance hooks (`claim-guard`,\n`test-tamper-guard`, `journey-coverage`): those *gate silently on static signals*; this\nis the human\u002Forchestrator-facing reviewer that *judges the running-app evidence and\nreturns a verdict*. See `docs\u002Fagentic-perfection\u002FAGENTIC-REPAIR-RELIABILITY.md`\n(NORTH-STAR REPAIR FLOW: \"fresh-context \u002F cross-model acceptance review over the real\nrunning app\").\n\n## When to run\n- Before claiming an app\u002Ffeature is done, works, or ships; before opening a PR.\n- At the end of a repair session — especially after an assertive \"fixed \u002F it works now\".\n- Whenever \"coverage is green\" is the only evidence offered that the journey works.\n\n## Procedure\n\n**1. Gather evidence (deterministic).** Run the helper **by its own path** — it ships\nbeside this SKILL.md, NOT inside the repository under review, and `--repo` (never the\nworking directory) is what selects that repository. Read `\u003CSKILL_DIR>` as the directory\nholding this file: `~\u002F.agents\u002Fskills\u002Facceptance-review` once installed,\n`.agents\u002Fskills\u002Facceptance-review` inside the toolchain repo. It collects the change\nset, the declared journeys, the artifact evidence, the reward-hack surface, and probes\nwhich cross-model CLIs are actually on PATH:\n\n```\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> [--base \u003CREF>] [--out \u003CDIR>]\n```\n\nBase defaults to the merge-base with a mainline (`origin\u002Fmain`→`main`→…→`HEAD~1`). It\nwrites `review-prompt.md` + `evidence.json` under `\u003CREPO>\u002F.acceptance-review\u002F` and prints\na routing line: which reviewer was chosen and why. The gathered evidence is the four\nexternal signals a fresh reviewer needs:\n- **Diff** (`git diff \u003Cbase>...HEAD`) — what actually changed.\n- **Declared journeys** — journey\u002FE2E spec files on disk (same globs as `journey-coverage`).\n- **Artifacts** — on-disk proof a journey ran (`job.done`, `*.mp4`, `playwright-report\u002F`,\n  `test-results\u002F`, screenshots, reports), newest-first, flagged fresh vs. stale relative\n  to the change.\n- **Reward-hack surface** — net-new `pragma:no-cover` \u002F `xfail` \u002F `skip` \u002F coverage-omit \u002F\n  marker-deselect \u002F CI-disable in the diff (where gaming hides).\n\n**2. Dispatch an INDEPENDENT reviewer** (prefer a different model; degrade gracefully).\nThe helper's routing (`--reviewer auto`, the default) picks in this order:\n\n| Reviewer | When | How |\n|---|---|---|\n| **Codex CLI** | `codex` on PATH | `--dispatch` runs `codex exec -s read-only -C \u003Crepo>` with the adversarial prompt (read-only: it may open the artifact, never mutate the repo). Different model → genuine cross-model verifier. |\n| **Gemini CLI** | `codex` absent, `gemini` present | `--dispatch` runs `gemini --approval-mode plan -o text -p \u003Cprompt>` (plan = read-only). |\n| **Fresh Claude subagent** | neither CLI present (or `--reviewer claude`) | The helper CANNOT spawn it (a Python process can't call the Task tool). YOU spawn a **fresh-context** subagent via the `Agent`\u002F`Task` tool (general-purpose or Explore) whose **only** input is `review-prompt.md` — do NOT paste your own reasoning; fresh context is the whole point (mirrors `fresh-context-verify`). |\n\nTo actually run a CLI reviewer, add `--dispatch` (wrap it with `longrun` \u002F\n`run_in_background` — a cross-model agentic review routinely exceeds 60s; default\ntimeout 240s). A timeout or an unparseable reply is treated as **BLOCK**, never a pass.\n\n```\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> --dispatch          # auto → codex\u002Fgemini\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> --reviewer gemini --dispatch --model gemini-2.5-pro\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> --reviewer claude   # prints the fresh-subagent instruction\n```\n\n**3. Read back the structured verdict.** Every reviewer must end with the machine-parsed\nblock the prompt mandates:\n\n```\nVERDICT: SHIP | BLOCK\nUNPROVEN_JOURNEYS: \u003Cspec paths with no post-change artifact, or NONE>\nGAPS: \u003Cconcrete runtime failures \u002F severed steps, or NONE>\nREWARD_HACK_FINDINGS: \u003Cexam-weakening diff changes, or NONE>\nDEMANDED_ARTIFACTS: \u003Cthe real output files required to flip BLOCK→SHIP>\nCONFIDENCE: \u003C0.0–1.0>\nRATIONALE: \u003C2–4 sentences; coverage \u002F \"tests pass\" forbidden as evidence>\n```\n\nRelay it to the user verbatim. **SHIP only if every declared journey has a real,\npost-change artifact proving the flow.** Zero declared journeys is itself a BLOCK-worthy\ngap (no held-out done-signal exists) — surface that.\n\n## The adversarial contract (baked into the prompt)\n1. Assume green-but-broken; for each journey, name the runtime step that could fail.\n2. **Demand the artifact** — the produced file a human could open, not a unit test.\n3. **Forbid trusting coverage \u002F \"all tests pass\"** — those are the gameable proxy.\n4. Inspect the reward-hack surface; treat net-new pragma\u002Fxfail\u002Fskip\u002Fomit as suspect.\n5. A journey spec with no post-change artifact is **UNPROVEN** (≠ passing) — list it.\n\n## Guardrails\n- **Never self-certify.** If `--reviewer` resolves to `claude`, the subagent must be\n  fresh-context; do not review your own work in the same window.\n- **Read-only.** CLI reviewers run sandboxed read-only \u002F plan mode; this skill never\n  edits app source (it judges, it doesn't fix).\n- **Fail toward BLOCK.** Missing evidence, a timeout, or an unparseable verdict is\n  inconclusive → treat as BLOCK, not a pass.\n- **Advisory exit by default.** The helper exits 0 unless `--strict` is set with a\n  dispatched BLOCK verdict (then exit 20) — for wiring into a gate.\n\n## Files\n```\nacceptance-review\u002F\n  SKILL.md                          — this skill\n  README.md                         — install path + rationale\n  scripts\u002Facceptance_review.py      — evidence gather + prompt build + route + dispatch (stdlib)\n  scripts\u002Ftest_acceptance_review.py — unit tests for the pure core\n```\n","# Acceptance Review — the cross-model \"does it actually work?\" verdict\n\nThe external-signal verifier for the \"done\" boundary. A capable model optimizes its\nown done-signal — coverage, \"tests pass\" — and stops, while the real product is\nbroken (dead buttons, \"shorts don't generate\", a severed A→B→C flow). The signal is\ngameable because the *same context that wrote the code certifies it*: intrinsic\nself-correction degrades (Huang, ICLR'24), verifier↔generator errors couple\n(\"Verification Mirage\"), and judges prefer their own output (Panickssery, NeurIPS'24).\nThis skill breaks that loop with **verifier separation** — an evaluator that did NOT\nwrite the code and, when possible, is a **different model** (VeriMAP \u002F MetaGPT), which\ndemands the real artifact and is forbidden from trusting coverage.\n\nIt complements the always-on functional-acceptance hooks (`claim-guard`,\n`test-tamper-guard`, `journey-coverage`): those *gate silently on static signals*; this\nis the human\u002Forchestrator-facing reviewer that *judges the running-app evidence and\nreturns a verdict*. See `docs\u002Fagentic-perfection\u002FAGENTIC-REPAIR-RELIABILITY.md`\n(NORTH-STAR REPAIR FLOW: \"fresh-context \u002F cross-model acceptance review over the real\nrunning app\").\n\n## When to run\n- Before claiming an app\u002Ffeature is done, works, or ships; before opening a PR.\n- At the end of a repair session — especially after an assertive \"fixed \u002F it works now\".\n- Whenever \"coverage is green\" is the only evidence offered that the journey works.\n\n## Procedure\n\n**1. Gather evidence (deterministic).** Run the helper **by its own path** — it ships\nbeside this SKILL.md, NOT inside the repository under review, and `--repo` (never the\nworking directory) is what selects that repository. Read `\u003CSKILL_DIR>` as the directory\nholding this file: `~\u002F.agents\u002Fskills\u002Facceptance-review` once installed,\n`.agents\u002Fskills\u002Facceptance-review` inside the toolchain repo. It collects the change\nset, the declared journeys, the artifact evidence, the reward-hack surface, and probes\nwhich cross-model CLIs are actually on PATH:\n\n```\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> [--base \u003CREF>] [--out \u003CDIR>]\n```\n\nBase defaults to the merge-base with a mainline (`origin\u002Fmain`→`main`→…→`HEAD~1`). It\nwrites `review-prompt.md` + `evidence.json` under `\u003CREPO>\u002F.acceptance-review\u002F` and prints\na routing line: which reviewer was chosen and why. The gathered evidence is the four\nexternal signals a fresh reviewer needs:\n- **Diff** (`git diff \u003Cbase>...HEAD`) — what actually changed.\n- **Declared journeys** — journey\u002FE2E spec files on disk (same globs as `journey-coverage`).\n- **Artifacts** — on-disk proof a journey ran (`job.done`, `*.mp4`, `playwright-report\u002F`,\n  `test-results\u002F`, screenshots, reports), newest-first, flagged fresh vs. stale relative\n  to the change.\n- **Reward-hack surface** — net-new `pragma:no-cover` \u002F `xfail` \u002F `skip` \u002F coverage-omit \u002F\n  marker-deselect \u002F CI-disable in the diff (where gaming hides).\n\n**2. Dispatch an INDEPENDENT reviewer** (prefer a different model; degrade gracefully).\nThe helper's routing (`--reviewer auto`, the default) picks in this order:\n\n| Reviewer | When | How |\n|---|---|---|\n| **Codex CLI** | `codex` on PATH | `--dispatch` runs `codex exec -s read-only -C \u003Crepo>` with the adversarial prompt (read-only: it may open the artifact, never mutate the repo). Different model → genuine cross-model verifier. |\n| **Gemini CLI** | `codex` absent, `gemini` present | `--dispatch` runs `gemini --approval-mode plan -o text -p \u003Cprompt>` (plan = read-only). |\n| **Fresh Claude subagent** | neither CLI present (or `--reviewer claude`) | The helper CANNOT spawn it (a Python process can't call the Task tool). YOU spawn a **fresh-context** subagent via the `Agent`\u002F`Task` tool (general-purpose or Explore) whose **only** input is `review-prompt.md` — do NOT paste your own reasoning; fresh context is the whole point (mirrors `fresh-context-verify`). |\n\nTo actually run a CLI reviewer, add `--dispatch` (wrap it with `longrun` \u002F\n`run_in_background` — a cross-model agentic review routinely exceeds 60s; default\ntimeout 240s). A timeout or an unparseable reply is treated as **BLOCK**, never a pass.\n\n```\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> --dispatch          # auto → codex\u002Fgemini\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> --reviewer gemini --dispatch --model gemini-2.5-pro\npython \u003CSKILL_DIR>\u002Fscripts\u002Facceptance_review.py --repo \u003CREPO> --reviewer claude   # prints the fresh-subagent instruction\n```\n\n**3. Read back the structured verdict.** Every reviewer must end with the machine-parsed\nblock the prompt mandates:\n\n```\nVERDICT: SHIP | BLOCK\nUNPROVEN_JOURNEYS: \u003Cspec paths with no post-change artifact, or NONE>\nGAPS: \u003Cconcrete runtime failures \u002F severed steps, or NONE>\nREWARD_HACK_FINDINGS: \u003Cexam-weakening diff changes, or NONE>\nDEMANDED_ARTIFACTS: \u003Cthe real output files required to flip BLOCK→SHIP>\nCONFIDENCE: \u003C0.0–1.0>\nRATIONALE: \u003C2–4 sentences; coverage \u002F \"tests pass\" forbidden as evidence>\n```\n\nRelay it to the user verbatim. **SHIP only if every declared journey has a real,\npost-change artifact proving the flow.** Zero declared journeys is itself a BLOCK-worthy\ngap (no held-out done-signal exists) — surface that.\n\n## The adversarial contract (baked into the prompt)\n1. Assume green-but-broken; for each journey, name the runtime step that could fail.\n2. **Demand the artifact** — the produced file a human could open, not a unit test.\n3. **Forbid trusting coverage \u002F \"all tests pass\"** — those are the gameable proxy.\n4. Inspect the reward-hack surface; treat net-new pragma\u002Fxfail\u002Fskip\u002Fomit as suspect.\n5. A journey spec with no post-change artifact is **UNPROVEN** (≠ passing) — list it.\n\n## Guardrails\n- **Never self-certify.** If `--reviewer` resolves to `claude`, the subagent must be\n  fresh-context; do not review your own work in the same window.\n- **Read-only.** CLI reviewers run sandboxed read-only \u002F plan mode; this skill never\n  edits app source (it judges, it doesn't fix).\n- **Fail toward BLOCK.** Missing evidence, a timeout, or an unparseable verdict is\n  inconclusive → treat as BLOCK, not a pass.\n- **Advisory exit by default.** The helper exits 0 unless `--strict` is set with a\n  dispatched BLOCK verdict (then exit 20) — for wiring into a gate.\n\n## Files\n```\nacceptance-review\u002F\n  SKILL.md                          — this skill\n  README.md                         — install path + rationale\n  scripts\u002Facceptance_review.py      — evidence gather + prompt build + route + dispatch (stdlib)\n  scripts\u002Ftest_acceptance_review.py — unit tests for the pure core\n```\n",{"name":5,"description":7},"main","a298c929b6cbd662e1fce54302ce520d5ceea1a5","skills\u002Facceptance-review","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain","https:\u002F\u002Fgithub.com\u002FPrekzursil","2026-11-02 14:30:57",{"source":8,"skillName":5,"commands":56,"prompt":66,"zipAvailable":67},[57,60,63],{"tool":58,"command":59},"npx","npx skills add Prekzursil\u002Fagent-skills-toolchain --skill acceptance-review",{"tool":61,"command":62},"bunx","bunx skills add Prekzursil\u002Fagent-skills-toolchain --skill acceptance-review",{"tool":64,"command":65},"pnpm","pnpm dlx skills add Prekzursil\u002Fagent-skills-toolchain --skill acceptance-review","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Facceptance-review\u002FSKILL.md",true,[69,75,79,83,87,91,95,99,103,107,111,115,119,123,127,131,135,141],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"README.md","REFERENCE","参考文档",3315,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Facceptance-review\u002FREADME.md",{"path":76,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":77,"sourceHtmlUrl":78},"references\u002Fauth-best-practices.md",6207,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fauth-best-practices.md",{"path":80,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":81,"sourceHtmlUrl":82},"references\u002Fsdk\u002Fazure-ai-contentsafety-java.md",1401,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-contentsafety-java.md",{"path":84,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":85,"sourceHtmlUrl":86},"references\u002Fsdk\u002Fazure-ai-contentsafety-py.md",1361,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-contentsafety-py.md",{"path":88,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":89,"sourceHtmlUrl":90},"references\u002Fsdk\u002Fazure-ai-contentsafety-ts.md",1413,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-contentsafety-ts.md",{"path":92,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":93,"sourceHtmlUrl":94},"references\u002Fsdk\u002Fazure-ai-document-intelligence-dotnet.md",1456,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-document-intelligence-dotnet.md",{"path":96,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":97,"sourceHtmlUrl":98},"references\u002Fsdk\u002Fazure-ai-document-intelligence-ts.md",1573,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-document-intelligence-ts.md",{"path":100,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":101,"sourceHtmlUrl":102},"references\u002Fsdk\u002Fazure-ai-openai-dotnet.md",1419,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-openai-dotnet.md",{"path":104,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":105,"sourceHtmlUrl":106},"references\u002Fsdk\u002Fazure-ai-transcription-py.md",1210,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-transcription-py.md",{"path":108,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":109,"sourceHtmlUrl":110},"references\u002Fsdk\u002Fazure-ai-translation-text-py.md",1345,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-translation-text-py.md",{"path":112,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":113,"sourceHtmlUrl":114},"references\u002Fsdk\u002Fazure-ai-translation-ts.md",1369,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-translation-ts.md",{"path":116,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":117,"sourceHtmlUrl":118},"references\u002Fsdk\u002Fazure-ai-vision-imageanalysis-java.md",1325,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-vision-imageanalysis-java.md",{"path":120,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":121,"sourceHtmlUrl":122},"references\u002Fsdk\u002Fazure-ai-vision-imageanalysis-py.md",1297,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-ai-vision-imageanalysis-py.md",{"path":124,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":125,"sourceHtmlUrl":126},"references\u002Fsdk\u002Fazure-search-documents-dotnet.md",1334,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-search-documents-dotnet.md",{"path":128,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":129,"sourceHtmlUrl":130},"references\u002Fsdk\u002Fazure-search-documents-py.md",1561,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-search-documents-py.md",{"path":132,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":133,"sourceHtmlUrl":134},"references\u002Fsdk\u002Fazure-search-documents-ts.md",1259,"https:\u002F\u002Fgithub.com\u002Fstrmt7\u002Fai_skills_collection_benchmarked\u002Fblob\u002Fmain\u002Fincluded\u002Fskills\u002Fby-category\u002Fcloud-azure-microsoft-sdks\u002Fofficial-vendor-reference\u002Fazure-ai\u002Freferences\u002Fsdk\u002Fazure-search-documents-ts.md",{"path":136,"resourceType":137,"resourceTypeDesc":138,"sizeBytes":139,"sourceHtmlUrl":140},"scripts\u002Facceptance_review.py","SCRIPT","脚本",24400,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Facceptance-review\u002Fscripts\u002Facceptance_review.py",{"path":142,"resourceType":137,"resourceTypeDesc":138,"sizeBytes":143,"sourceHtmlUrl":144},"scripts\u002Ftest_acceptance_review.py",6627,"https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Fskills\u002Facceptance-review\u002Fscripts\u002Ftest_acceptance_review.py",[146,167,196,222,236,252],{"skillNo":147,"name":148,"nameZh":6,"description":149,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":150,"categories":151,"keywords":152,"publishedAt":44},"01M6DF7D7478R55P46REYH4QAC","clarify-requirements","Resolve genuine pre-execution ambiguity by scanning repo context, generating 1-6 grounded options, and asking 1-3 focused questions via AskUserQuestion. NOT for mid-task use (injection risk); NOT when the goal is already unambiguous; use to disambiguate vague prompts before any file edits.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fworkspace-hooks\u002Fskills\u002Fclarify-requirements\u002FSKILL.md",[],[153,154,155,156,157,158,159,160,161,36,162,163,164,165,166],"clarify requirements","clarify","requirements","risk","options","task","prompts","goal","scanning","context","repo","injection","execution","file",{"skillNo":168,"name":169,"nameZh":6,"description":170,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":171,"categories":172,"keywords":176,"publishedAt":44},"01M6DF79CA4BHJS66J8EZDBPWJ","openmm-md","Run FREE, open-source, Windows-native GPU molecular dynamics via the openmm-md MCP (OpenMM + CUDA on the local NVIDIA GPU). Use for protein \u002F protein-ligand MD when you want a free stack or are on Windows without a Schrodinger license. Covers the full workflow (prepare, solvate, minimize, equilibrate, production, analyze, mmgbsa) AND an honest decision guide vs Schrodinger Desmond, including how to run Desmond legitimately on Linux or a cloud GPU with a real license. This is a deliberately LESSER free alternative and does NOT replace the schrodinger MCP tools. Do NOT use it to pick a free tool when a validated FEP+, WaterMap, IFD-MD, or Prime MM-GBSA result is actually required.","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Fmcp\u002Fopenmm-md-mcp\u002Fskill\u002Fopenmm-md\u002FSKILL.md",[173],{"categoryNo":174,"categoryName":175,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000023","Skill与智能体开发",[177,178,179,180,181,182,183,184,185,186,187,188,189,190,36,191,192,193,194,195],"openmm md","how","decision","pick","linux","analyze","tool","stack","open","NVIDIA","gpu","license","workflow","windows","free","molecular dynamics","local","do","MCP",{"skillNo":197,"name":198,"nameZh":6,"description":199,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":200,"categories":201,"keywords":208,"publishedAt":44},"01M6DF76C1MG85VEWP8VZCBNMN","visual-review","Take screenshots of web pages and UI using Playwright for visual review and iteration","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fvisual-review\u002FSKILL.md",[202,205],{"categoryNo":203,"categoryName":204,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000003","演示文稿",{"categoryNo":206,"categoryName":207,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000008","测试与质量",[24,209,210,211,212,213,214,215,216,217,178,218,219,220,221],"visual","visual review","pages","web","false","iteration","take","check","slides","ui","screenshot","playwright","screenshots",{"skillNo":223,"name":224,"nameZh":6,"description":225,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":226,"categories":227,"keywords":228,"publishedAt":44},"01M6DF73AQVCBZKEN0YXEX9FXP","status","Diagnostic status report — shows metaswarm installation state, project setup, and potential issues","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstatus\u002FSKILL.md",[],[224,229,230,231,232,233,234,235],"setup","report","project","installation","project setup","issues","state",{"skillNo":237,"name":238,"nameZh":6,"description":239,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":240,"categories":241,"keywords":243,"publishedAt":44},"01M6DF6X6C2XB0P6GX76P7TTYV","start","Use when starting work on any task, when the user mentions metaswarm, or when the user wants to begin tracked development work","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fstart\u002FSKILL.md",[242],{"categoryNo":174,"categoryName":175,"taxonomyType":21,"taxonomyTypeDesc":22},[238,244,158,245,246,247,36,248,249,250,251],"start issue","issue","ready","work on issue","user","work","work on","development",{"skillNo":253,"name":229,"nameZh":6,"description":254,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":255,"categories":256,"keywords":257,"publishedAt":44},"01M6DF6QMJ9AJKFKHHKRT005PQ","Interactive project setup — detects your project, configures metaswarm, writes project-local files","https:\u002F\u002Fgithub.com\u002FPrekzursil\u002Fagent-skills-toolchain\u002Fblob\u002Fmain\u002Ftools\u002Feditor\u002Fmetaswarm\u002Fskills\u002Fsetup\u002FSKILL.md",[],[229,231,193,258,233,259],"files","interactive"]