[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7D2QHC8WYXYEXSXNATQ0WD2":3,"skill-same:01M7D2QHC8WYXYEXSXNATQ0WD2":49},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":30,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":33,"ownerDescription":6,"updatedAt":34,"install":35,"resources":48},"01M7D2QHC8WYXYEXSXNATQ0WD2","explain",null,"Human-started command: it runs only when the human's message begins with `\u002Fexplain`. On any other request do not load or follow it; tell the human to type that command. Explains how and why existing behavior works, from repository and knowledgebase evidence, with every claim traced to a source, marked as inference, or marked unknown. Use when a human asks how or why something in the project works the way it does. Not for \"should we change it\", which is a judgment, and not for fixing a failure or planning a change.","Pibomeister\u002Fagent-kit","Pibomeister","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F11633791?v=4","01M7D2NBHEKARMHN9MPV2J80XD","Eduardo Picazo","01M7D2NBJ3Z33SGSYCZ0KBD5DE",0,"MIT","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002F.cursor\u002Fskills\u002Fexplain\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: explain\ndescription: \"Human-started command: it runs only when the human's message begins with `\u002Fexplain`. On any other request do not load or follow it; tell the human to type that command. Explains how and why existing behavior works, from repository and knowledgebase evidence, with every claim traced to a source, marked as inference, or marked unknown. Use when a human asks how or why something in the project works the way it does. Not for \\\"should we change it\\\", which is a judgment, and not for fixing a failure or planning a change.\"\nlicense: MIT\nmetadata:\n  ak_catalog_id: explain\n  ak:\n    mode: manual\n---\n\nExplain how and why existing behavior works from repository and KB evidence. A description is not\na recommendation or an implementation plan.\n\n## When to use\n\n- A human asks how a part of the project works, or why it works the way it does, and wants the\n  answer grounded in the code and the project's records.\n- A human is onboarding to an area and needs a description that separates what the evidence shows\n  from what is inferred.\n\n## Not for\n\n- \"Should we change X?\" That is a judgment on adoption, and it belongs to `pov`, which the human\n  starts. Explaining a historical choice is not endorsing it today.\n- Capturing what was learned as durable project knowledge. That is `compound`, and an explanation\n  does not authorize writing to the knowledgebase.\n- Generating alternatives or scoping an implementation. That is `ideate`, `super-align` or\n  `super-bound`, each started by the human.\n- A reported failure to find and fix. That is `diagnose`. A factual account of current behavior\n  stays here.\n- Re-pitching the agent's own last message in plainer words. That is `wait-what`.\n\n## Authority\n\nAuthority: `explicit`. A human starts it with `\u002Fexplain`. No phase operation exposes it\n([`policies\u002Finvocation.yaml`](..\u002F..\u002Freferences\u002Fshared\u002Fpolicies\u002Finvocation.yaml) lists it among the user-invoked skills that expose none), so no\ncontroller can start it under a grant. Started any other way, it stops at step 1 and names the\ncommand.\n\n## Inputs\n\n- The subject: a behavior, a file, a symbol or a past decision. Unclear: see step 2.\n- The repository at a readable revision. Unreadable: `needs-input`.\n- Knowledgebase context through the adapter's `readContext`: decision records, `concept` and\n  `system` pages, the glossary. Optional. Unavailable: the explanation says so and rests on the\n  repository alone.\n- The intended readers, when the request names them. Depth follows the request, never the caller's\n  identity alone.\n\n## Workflow\n\n1. **Check how this run was started**, before any other step and before any tool call. It is started\n   only when the human's message begins with `\u002Fexplain`; no grant starts it. A request in prose\n   is not a start, even when it names this skill or the command. Otherwise, stop here: make no tool\n   call, say that this command is human-started, and give the human the line to type, `\u002Fexplain`\n   and their request.\n2. **Resolve the subject.** Resolve discoverable facts before asking. Ask only when the missing\n   information changes the answer. With no one to ask, return the unresolved question and what it\n   changes, rather than guessing.\n3. **Gather evidence.** Read the implementation, not only its call sites. Read the decision records\n   for the reasons. Record each source by path and line range, or by record id.\n4. **Classify each claim.** Every claim is traced to a source, marked `inference` with what it rests\n   on, or marked `unknown`. A function call does not establish guarantees about its uninspected\n   implementation. Code shows behavior, not intent. A search that found no reason does not prove\n   there was none. An outside concept with no source is labelled unverified.\n5. **Write the explanation.** Load [the prose-quality reference pack](..\u002F..\u002Freferences\u002Fshared\u002Freferences\u002Fprose-quality\u002FREFERENCE.md)\n   and write to it. Fit the depth to the readers. When the evidence is wider than the answer,\n   say what was selected and what was left out.\n6. **Check before delivery.** Check every factual claim against its source. Remove an unsupported\n   claim or mark its uncertainty where it appears.\n7. **Deliver and stop.** Return the explanation, its sources and the open questions in the session.\n   End there: no recommendation, no plan. A follow-up \"so what should we do?\" is answered by naming\n   `pov` as the human's to start.\n\n## Hard gates\n\nGate: no recommendation and no implementation plan. A description that ends in a verdict or a list\nof changes is refused, however the request is worded.\n\nGate: no claim about code that was not read. Where only a call site is available, the behavior\nbehind it is `unknown`, and the explanation says so.\n\nGate: no claim delivered as fact without a source. It is traced, marked `inference`, or marked\n`unknown`.\n\nGate: nothing durable is written. The explanation is delivered in the session; publishing it is a\nseparate action and not this skill's.\n\nGate: no invented answer to an open question. With no one to ask, the question and its consequence\nare returned.\n\n| The thought | Why it is wrong | Do this instead |\n|---|---|---|\n| \"The function is called `retryWithBackoff`, so it retries with backoff.\" | A name and a call site do not establish what an uninspected implementation guarantees. | Read the implementation, or mark the behavior `unknown`. |\n| \"They will want to know what to do next, so I'll add a suggestion at the end.\" | A description is not a recommendation; the suggestion is a judgment nobody asked this skill to make. | End with the evidence and the open questions. Name `pov` if a judgment is wanted. |\n| \"The commit message gives no reason, so there wasn't one.\" | Code shows behavior, not intent, and an empty search is not evidence of absence. | Mark the reason `unknown` and say where you looked. |\n| \"It was chosen deliberately, so it is still the right choice.\" | Explaining a historical choice is not endorsing it today. | Report the reason as it was recorded, with its date and source. |\n| \"This explanation is good; I'll save it as a doc for next time.\" | An explanation does not authorize a durable write, and project knowledge has its own owner. | Deliver it in the session. Name `compound` if it should be kept. |\n\n## Outputs\n\n- The explanation, returned in the session: each claim with its source, its `inference` marker or\n  its `unknown` marker, plus the material open questions. It is not published: no knowledgebase\n  operation is called (`docs\u002Fdecisions\u002F0001-kb-document-vocabulary.md` §3 lists no durable output\n  for this skill).\n- With no one to ask, the unresolved question and what its answer would change.\n\n## Side effects\n\nNone. The explanation is returned in the session, and nothing is written to the repository, the\nknowledgebase or the run's artifact store.\n\n## Stop conditions\n\n- `complete` — the explanation is delivered, with every claim traced, inferred or unknown.\n- `needs-input` — started without an explicit `\u002Fexplain`; or the repository is unreadable; or an\n  open question changes the answer and no one can answer it, returned with its consequence.\n\n## Limits\n\n- Claims without a source, an `inference` marker or an `unknown` marker: 0 (gate).\n- Clarifying questions: only those whose answer changes the explanation (guidance).\n","Explain how and why existing behavior works from repository and KB evidence. A description is not\na recommendation or an implementation plan.\n\n## When to use\n\n- A human asks how a part of the project works, or why it works the way it does, and wants the\n  answer grounded in the code and the project's records.\n- A human is onboarding to an area and needs a description that separates what the evidence shows\n  from what is inferred.\n\n## Not for\n\n- \"Should we change X?\" That is a judgment on adoption, and it belongs to `pov`, which the human\n  starts. Explaining a historical choice is not endorsing it today.\n- Capturing what was learned as durable project knowledge. That is `compound`, and an explanation\n  does not authorize writing to the knowledgebase.\n- Generating alternatives or scoping an implementation. That is `ideate`, `super-align` or\n  `super-bound`, each started by the human.\n- A reported failure to find and fix. That is `diagnose`. A factual account of current behavior\n  stays here.\n- Re-pitching the agent's own last message in plainer words. That is `wait-what`.\n\n## Authority\n\nAuthority: `explicit`. A human starts it with `\u002Fexplain`. No phase operation exposes it\n([`policies\u002Finvocation.yaml`](..\u002F..\u002Freferences\u002Fshared\u002Fpolicies\u002Finvocation.yaml) lists it among the user-invoked skills that expose none), so no\ncontroller can start it under a grant. Started any other way, it stops at step 1 and names the\ncommand.\n\n## Inputs\n\n- The subject: a behavior, a file, a symbol or a past decision. Unclear: see step 2.\n- The repository at a readable revision. Unreadable: `needs-input`.\n- Knowledgebase context through the adapter's `readContext`: decision records, `concept` and\n  `system` pages, the glossary. Optional. Unavailable: the explanation says so and rests on the\n  repository alone.\n- The intended readers, when the request names them. Depth follows the request, never the caller's\n  identity alone.\n\n## Workflow\n\n1. **Check how this run was started**, before any other step and before any tool call. It is started\n   only when the human's message begins with `\u002Fexplain`; no grant starts it. A request in prose\n   is not a start, even when it names this skill or the command. Otherwise, stop here: make no tool\n   call, say that this command is human-started, and give the human the line to type, `\u002Fexplain`\n   and their request.\n2. **Resolve the subject.** Resolve discoverable facts before asking. Ask only when the missing\n   information changes the answer. With no one to ask, return the unresolved question and what it\n   changes, rather than guessing.\n3. **Gather evidence.** Read the implementation, not only its call sites. Read the decision records\n   for the reasons. Record each source by path and line range, or by record id.\n4. **Classify each claim.** Every claim is traced to a source, marked `inference` with what it rests\n   on, or marked `unknown`. A function call does not establish guarantees about its uninspected\n   implementation. Code shows behavior, not intent. A search that found no reason does not prove\n   there was none. An outside concept with no source is labelled unverified.\n5. **Write the explanation.** Load [the prose-quality reference pack](..\u002F..\u002Freferences\u002Fshared\u002Freferences\u002Fprose-quality\u002FREFERENCE.md)\n   and write to it. Fit the depth to the readers. When the evidence is wider than the answer,\n   say what was selected and what was left out.\n6. **Check before delivery.** Check every factual claim against its source. Remove an unsupported\n   claim or mark its uncertainty where it appears.\n7. **Deliver and stop.** Return the explanation, its sources and the open questions in the session.\n   End there: no recommendation, no plan. A follow-up \"so what should we do?\" is answered by naming\n   `pov` as the human's to start.\n\n## Hard gates\n\nGate: no recommendation and no implementation plan. A description that ends in a verdict or a list\nof changes is refused, however the request is worded.\n\nGate: no claim about code that was not read. Where only a call site is available, the behavior\nbehind it is `unknown`, and the explanation says so.\n\nGate: no claim delivered as fact without a source. It is traced, marked `inference`, or marked\n`unknown`.\n\nGate: nothing durable is written. The explanation is delivered in the session; publishing it is a\nseparate action and not this skill's.\n\nGate: no invented answer to an open question. With no one to ask, the question and its consequence\nare returned.\n\n| The thought | Why it is wrong | Do this instead |\n|---|---|---|\n| \"The function is called `retryWithBackoff`, so it retries with backoff.\" | A name and a call site do not establish what an uninspected implementation guarantees. | Read the implementation, or mark the behavior `unknown`. |\n| \"They will want to know what to do next, so I'll add a suggestion at the end.\" | A description is not a recommendation; the suggestion is a judgment nobody asked this skill to make. | End with the evidence and the open questions. Name `pov` if a judgment is wanted. |\n| \"The commit message gives no reason, so there wasn't one.\" | Code shows behavior, not intent, and an empty search is not evidence of absence. | Mark the reason `unknown` and say where you looked. |\n| \"It was chosen deliberately, so it is still the right choice.\" | Explaining a historical choice is not endorsing it today. | Report the reason as it was recorded, with its date and source. |\n| \"This explanation is good; I'll save it as a doc for next time.\" | An explanation does not authorize a durable write, and project knowledge has its own owner. | Deliver it in the session. Name `compound` if it should be kept. |\n\n## Outputs\n\n- The explanation, returned in the session: each claim with its source, its `inference` marker or\n  its `unknown` marker, plus the material open questions. It is not published: no knowledgebase\n  operation is called (`docs\u002Fdecisions\u002F0001-kb-document-vocabulary.md` §3 lists no durable output\n  for this skill).\n- With no one to ask, the unresolved question and what its answer would change.\n\n## Side effects\n\nNone. The explanation is returned in the session, and nothing is written to the repository, the\nknowledgebase or the run's artifact store.\n\n## Stop conditions\n\n- `complete` — the explanation is delivered, with every claim traced, inferred or unknown.\n- `needs-input` — started without an explicit `\u002Fexplain`; or the repository is unreadable; or an\n  open question changes the answer and no one can answer it, returned with its consequence.\n\n## Limits\n\n- Claims without a source, an `inference` marker or an `unknown` marker: 0 (gate).\n- Clarifying questions: only those whose answer changes the explanation (guidance).\n",{"name":5,"license":15,"metadata":24,"description":7},{"ak":25,"ak_catalog_id":5},{"mode":26},"manual","main","e0232821c4abbcd4338d6018c20fb70f1b74336b",".cursor\u002Fskills\u002Fexplain",3,"https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit","https:\u002F\u002Fgithub.com\u002FPibomeister","www.eduardopicazoenriquez.com","2026-11-15 00:56:58",{"source":8,"skillName":5,"commands":36,"prompt":46,"zipAvailable":47},[37,40,43],{"tool":38,"command":39},"npx","npx skills add Pibomeister\u002Fagent-kit --skill explain",{"tool":41,"command":42},"bunx","bunx skills add Pibomeister\u002Fagent-kit --skill explain",{"tool":44,"command":45},"pnpm","pnpm dlx skills add Pibomeister\u002Fagent-kit --skill explain","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002F.cursor\u002Fskills\u002Fexplain\u002FSKILL.md",true,[],[50,57,62,67,72,77],{"skillNo":51,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":54,"categories":55,"keywords":56,"publishedAt":19},"01M7D3JZP2026RGY6XX1MC3NQQ","alpha","Human-started command: it runs only when the human's message begins with `\u002Fak:alpha`. On any other request do not load or follow it; tell the human to type that command. Runs the alpha workflow when a human asks for it.","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fvalid\u002Fminimal\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":58,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":61,"publishedAt":19},"01M7D3JA593Q3RBYJXKVZ37Y2K","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F25-article-tier-outside-vocabulary\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":63,"name":52,"nameZh":6,"description":53,"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},"01M7D3HQDDAZTAPGWCPGDPNVBQ","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F24-article-without-triggers\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":68,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":69,"categories":70,"keywords":71,"publishedAt":19},"01M7D3HK074TK3ZWK1ZZNQVPB0","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F23-enforced-by-without-predicate\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":73,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7D3HGHQM5C5TBSXBCE2WFHW","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F22-article-without-source-span\u002Fskills\u002Falpha\u002FSKILL.md",[],[],{"skillNo":78,"name":52,"nameZh":6,"description":53,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M7D3HDV6HMK2ZHRM0P368KNT","https:\u002F\u002Fgithub.com\u002FPibomeister\u002Fagent-kit\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Finvalid\u002F21-fragment-redeclares-id\u002Fskills\u002Falpha\u002FSKILL.md",[],[]]