[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7BV55MPCQBET9K4WN42RGP2":3,"skill-same:01M7BV55MPCQBET9K4WN42RGP2":93},{"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":25,"commitSha":26,"skillDirectory":27,"forksCount":28,"repositoryHtmlUrl":29,"repositoryHomepageUrl":30,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":32,"updatedAt":33,"install":34,"resources":46},"01M7BV55MPCQBET9K4WN42RGP2","comet-any",null,"通过 Comet Creator 创建或升级 Comet Classic workflow Skill。不用于一般 Skill 的编写、整理或评审。","cocoyes\u002Fzhizhi-agent-runtime","cocoyes","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F24385018?v=4","01M7BV53TBAA96RHJNP4TWH9A0","ApacheMoy","01M7BV53V1YC1R42809X7HEZ69",184,"MIT","https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: comet-any\ndescription: \"通过 Comet Creator 创建或升级 Comet Classic workflow Skill。不用于一般 Skill 的编写、整理或评审。\"\ndisable-model-invocation: true\n---\n\n# Comet Any - Skill Creator\n\n`\u002Fcomet-any` 是 Comet 的 Skill 创建向导。用户只需要描述想要的工作流；本 Skill 负责读取真实 Skill、提出方案、等待确认、生成可验证的 Comet-native Skill Bundle，并通过内部 CLI 完成 eval、review、publish readiness 和安装预览。\n\n普通用户第一层只看到三种起点：\n\n- `基于 \u002Fcomet-classic 的五阶段定制`：覆盖 `open \u002F design \u002F build \u002F verify \u002F archive` 五阶段的 Skill 编排，但不修改 `\u002Fcomet-classic` 永久入口本身。\n- `创建全新 workflow Skill`：从目标和候选 Skills 生成新的 `workflow-kernel`。\n- `整理已有 Skill`：读取已有 Skill，补齐 Workflow Node、Skill Binding、Output Schema、Guardrail、Handoff、eval 和 readiness。\n\n后端 Bundle、Factory、composition 仍是内部审计词；不要把它们作为普通用户的第一屏概念。\n\n## 核心模型\n\n所有路径都必须编译到同一种 Workflow Contract：\n\n- `Workflow Node`：流程中的可恢复节点，例如 `open`、`design`、`plan`、`execute`、`subagent-execute`、`review`、`verify`、`archive`。\n- `Node Responsibility`：该 Node 在 Agent workflow 中承担的职责，用来解释它为什么存在、需要产出什么、能否替换。\n- `Skill Binding`：某个 Node 的实现 Skill 或辅助 Skill。\n- `Required Skill Call`：要求 Node 内必须调用某个 Skill，不替换 Node implementation。例如 `execute` 和 `subagent-execute` 必须调用 `elementui`，`review` 必须调用 `whitebox-code-standard`。\n- `Output Schema`：Node 必须产出的文件、状态或 evidence。Output Schema 必须挂到具体 Workflow Node 才算生效；只定义在 `workflow.outputSchemas` 里不会触发 guard、eval 或 readiness。脚本、eval、readiness 只依赖挂到 Node 的 Output Schema，不依赖 Skill 名称。\n- `Guardrail`：阻断或放行 Node 推进的检查。\n- `Handoff`：子代理或跨 Node 交接时必须带回的 evidence。\n- `workflow-protocol.json`：生成包的唯一运行事实源，kind 为 `comet-five-phase-overlay` 或 `workflow-kernel`。\n\n## 受保护边界\n\n`comet-five-phase-overlay` 保留 Comet Classic 五阶段主流程和 `.comet.yaml` 状态语义。普通模式下：\n\n- `comet-five-phase-overlay` 的主状态只来自 Classic layout resolver 绑定的 `\u003Cclassic-change-dir>\u002F.comet.yaml`；没有 active change 或多个 active changes 时必须阻塞并请用户选择。\n- 不得创建 `.comet\u002Fruns\u002F\u003Cworkflow>\u002Fstate.json` 作为 Comet overlay 主状态。Bundle 草稿、eval evidence 和 publish readiness 可以有自己的证据文件，但不能替代 `.comet.yaml`。\n- `control` Node 不允许 override：`open`、`execute`、`verify`、`archive`。\n- `producer` Node 可以 override：`design`、`plan`，但必须满足对应 Output Schema。\n- `handoff` 和 `guardrail` Node 可以 require \u002F augment。\n- 用户坚持替换 control Node 时，改走高级 `workflow-kernel`，并要求重新声明 state、Output Schema 和 Guardrail。\n- 所有 Node 都必须用 responsibility 说明职责，不使用内部坐标作为用户理解流程的方式。\n\n## 工作步骤\n\n1. 恢复现有状态：先运行 `comet creator guide --project . --json`，展示恢复摘要和下一步。\n2. 读取项目偏好：读取 `.comet\u002Fskill-preferences.yaml`，用 `comet creator candidates --json` 发现真实本地 Skill，再用 `comet skill show \u003Cname> --json` 读取候选的真实内容与 hash。不得只按名字推测能力。\n3. 生成方案：把用户目标表达为 Workflow Nodes、Skill Bindings、Output Schemas、Guardrails、Handoffs 和 Evidence。\n4. 展示确认页：说明每个 Node 的职责、绑定 Skill、Required Skill Call、Output Schema、可执行披露和 readiness 影响。确认页必须为每个新增 binding 或 schema 显示 enforcement：`guarded`、`handoff-guarded`、`evidence-only` 或 `advisory`。\n5. 等待用户确认：未确认前不得写 Bundle draft；存在 missing \u002F ambiguous Skill 时必须暂停。\n6. 初始化后端状态：确认后调用 `comet creator init \u003Cname> --file \u003Cplan.json> --confirmed-proposal --json`。\n7. 运行创作管线并生成 Bundle：先运行 `comet creator authoring-plan \u003Cname> --depth quick|full --json` 取得 lane DAG。按 DAG 派发 lane——wave1（`script`、`reference`、`pause-points`）并发派发，wave2（`workflow-entry`、`skill-core`）在 script 契约之后并发派发，`skill-review` 作为汇聚 barrier。每个 lane 的产出用 `comet creator authoring-record \u003Cname> --lane \u003Cid> --file \u003Cout.json> --json` 记录（经 schema 校验；BLOCKED\u002FNEEDS_CONTEXT 会被拒绝）。随后运行 `comet creator generate \u003Cname> --json`：把记录的内容叶子草稿（entry\u002Fnode SKILL.md、decision-points、recovery）合并进包，而确定性脊梁（protocol\u002Fscripts\u002Fmanifest）保持模板化，并渲染真实审查证据。产出 entry Skill、Node Skills、`reference\u002Fworkflow-protocol.json`、六个 scripts、rules、hooks 与 `comet\u002Feval.yaml`。\n8. 验证：展示 quick\u002Ffull eval 工作量，运行或记录当前 draft hash 的 eval evidence；失败、skip 或证据 hash 过期时不得进入 ready。\n9. Review \u002F readiness：读取 `comet publish review \u003Cname> --platform \u003Creference-platform> --json`，展示 `Readiness:`、`Blockers:`、`Warnings:`、`Evidence:`。\n10. Publish \u002F install preview：人工批准后才能 publish；安装前必须先运行 preview，并展示 `No files were written`。\n\n## 方案示例\n\n组件库和白盒审查场景应生成类似 plan：\n\n```json\n{\n  \"goal\": \"基于 \u002Fcomet-classic 的五阶段定制，要求组件库和白盒审查。\",\n  \"skillCreatorIntent\": \"customize-comet\",\n  \"workflow\": {\n    \"kind\": \"comet-five-phase-overlay\",\n    \"name\": \"team-comet\",\n    \"goal\": \"要求组件库和白盒审查。\",\n    \"nodes\": {\n      \"execute\": {\n        \"requiredSkillCalls\": [\n          {\n            \"skill\": \"elementui\",\n            \"reason\": \"Use project component library during direct implementation.\"\n          }\n        ]\n      },\n      \"subagent-execute\": {\n        \"requiredSkillCalls\": [\n          {\n            \"skill\": \"elementui\",\n            \"scope\": \"handoff\"\n          }\n        ]\n      },\n      \"review\": {\n        \"requiredSkillCalls\": [\n          {\n            \"skill\": \"whitebox-code-standard\",\n            \"scope\": \"review\"\n          }\n        ]\n      }\n    }\n  }\n}\n```\n\n## 硬性规则\n\n- 必须先展示方案确认页，再生成。\n- 确认页必须为每个新增 binding 或 schema 显示 enforcement：`guarded`、`handoff-guarded`、`evidence-only` 或 `advisory`。\n- Required Skill Call 不替换 Node implementation。\n- producer override 必须声明 `satisfies` 的 Output Schema。\n- Output Schema 必须挂到具体 Workflow Node 才算生效；只定义在 `workflow.outputSchemas` 里不会触发 guard、eval 或 readiness。\n- control Node 普通模式不得 override。\n- eval、review、publish readiness 必须读取同一份 `workflow-protocol.json`。\n- readiness blockers 必须阻止 publish：缺少当前 draft hash 的 eval evidence、人工 approval、required capability 或 executable disclosure 任一项都不能进入 ready。\n- 子代理 Handoff 必须要求子代理加载 Required Skill Call 并回传 evidence。\n- 脚本只读取 protocol 和 state，不把 Skill 名称当成校验依据。\n- 安装前必须询问用户，不得自动安装。\n\n## 参考资料\n\n- `comet-any\u002Freference\u002Fauthoring-protocol.json`\n- `comet-any\u002Freference\u002Fauthored-zone-example.md`\n- `comet-any\u002Freference\u002Fbundle-authoring.md`\n- `comet-any\u002Freference\u002Fauthoring-subagents.md`\n- `comet-any\u002Freference\u002Feval-provider.md`\n","# Comet Any - Skill Creator\n\n`\u002Fcomet-any` 是 Comet 的 Skill 创建向导。用户只需要描述想要的工作流；本 Skill 负责读取真实 Skill、提出方案、等待确认、生成可验证的 Comet-native Skill Bundle，并通过内部 CLI 完成 eval、review、publish readiness 和安装预览。\n\n普通用户第一层只看到三种起点：\n\n- `基于 \u002Fcomet-classic 的五阶段定制`：覆盖 `open \u002F design \u002F build \u002F verify \u002F archive` 五阶段的 Skill 编排，但不修改 `\u002Fcomet-classic` 永久入口本身。\n- `创建全新 workflow Skill`：从目标和候选 Skills 生成新的 `workflow-kernel`。\n- `整理已有 Skill`：读取已有 Skill，补齐 Workflow Node、Skill Binding、Output Schema、Guardrail、Handoff、eval 和 readiness。\n\n后端 Bundle、Factory、composition 仍是内部审计词；不要把它们作为普通用户的第一屏概念。\n\n## 核心模型\n\n所有路径都必须编译到同一种 Workflow Contract：\n\n- `Workflow Node`：流程中的可恢复节点，例如 `open`、`design`、`plan`、`execute`、`subagent-execute`、`review`、`verify`、`archive`。\n- `Node Responsibility`：该 Node 在 Agent workflow 中承担的职责，用来解释它为什么存在、需要产出什么、能否替换。\n- `Skill Binding`：某个 Node 的实现 Skill 或辅助 Skill。\n- `Required Skill Call`：要求 Node 内必须调用某个 Skill，不替换 Node implementation。例如 `execute` 和 `subagent-execute` 必须调用 `elementui`，`review` 必须调用 `whitebox-code-standard`。\n- `Output Schema`：Node 必须产出的文件、状态或 evidence。Output Schema 必须挂到具体 Workflow Node 才算生效；只定义在 `workflow.outputSchemas` 里不会触发 guard、eval 或 readiness。脚本、eval、readiness 只依赖挂到 Node 的 Output Schema，不依赖 Skill 名称。\n- `Guardrail`：阻断或放行 Node 推进的检查。\n- `Handoff`：子代理或跨 Node 交接时必须带回的 evidence。\n- `workflow-protocol.json`：生成包的唯一运行事实源，kind 为 `comet-five-phase-overlay` 或 `workflow-kernel`。\n\n## 受保护边界\n\n`comet-five-phase-overlay` 保留 Comet Classic 五阶段主流程和 `.comet.yaml` 状态语义。普通模式下：\n\n- `comet-five-phase-overlay` 的主状态只来自 Classic layout resolver 绑定的 `\u003Cclassic-change-dir>\u002F.comet.yaml`；没有 active change 或多个 active changes 时必须阻塞并请用户选择。\n- 不得创建 `.comet\u002Fruns\u002F\u003Cworkflow>\u002Fstate.json` 作为 Comet overlay 主状态。Bundle 草稿、eval evidence 和 publish readiness 可以有自己的证据文件，但不能替代 `.comet.yaml`。\n- `control` Node 不允许 override：`open`、`execute`、`verify`、`archive`。\n- `producer` Node 可以 override：`design`、`plan`，但必须满足对应 Output Schema。\n- `handoff` 和 `guardrail` Node 可以 require \u002F augment。\n- 用户坚持替换 control Node 时，改走高级 `workflow-kernel`，并要求重新声明 state、Output Schema 和 Guardrail。\n- 所有 Node 都必须用 responsibility 说明职责，不使用内部坐标作为用户理解流程的方式。\n\n## 工作步骤\n\n1. 恢复现有状态：先运行 `comet creator guide --project . --json`，展示恢复摘要和下一步。\n2. 读取项目偏好：读取 `.comet\u002Fskill-preferences.yaml`，用 `comet creator candidates --json` 发现真实本地 Skill，再用 `comet skill show \u003Cname> --json` 读取候选的真实内容与 hash。不得只按名字推测能力。\n3. 生成方案：把用户目标表达为 Workflow Nodes、Skill Bindings、Output Schemas、Guardrails、Handoffs 和 Evidence。\n4. 展示确认页：说明每个 Node 的职责、绑定 Skill、Required Skill Call、Output Schema、可执行披露和 readiness 影响。确认页必须为每个新增 binding 或 schema 显示 enforcement：`guarded`、`handoff-guarded`、`evidence-only` 或 `advisory`。\n5. 等待用户确认：未确认前不得写 Bundle draft；存在 missing \u002F ambiguous Skill 时必须暂停。\n6. 初始化后端状态：确认后调用 `comet creator init \u003Cname> --file \u003Cplan.json> --confirmed-proposal --json`。\n7. 运行创作管线并生成 Bundle：先运行 `comet creator authoring-plan \u003Cname> --depth quick|full --json` 取得 lane DAG。按 DAG 派发 lane——wave1（`script`、`reference`、`pause-points`）并发派发，wave2（`workflow-entry`、`skill-core`）在 script 契约之后并发派发，`skill-review` 作为汇聚 barrier。每个 lane 的产出用 `comet creator authoring-record \u003Cname> --lane \u003Cid> --file \u003Cout.json> --json` 记录（经 schema 校验；BLOCKED\u002FNEEDS_CONTEXT 会被拒绝）。随后运行 `comet creator generate \u003Cname> --json`：把记录的内容叶子草稿（entry\u002Fnode SKILL.md、decision-points、recovery）合并进包，而确定性脊梁（protocol\u002Fscripts\u002Fmanifest）保持模板化，并渲染真实审查证据。产出 entry Skill、Node Skills、`reference\u002Fworkflow-protocol.json`、六个 scripts、rules、hooks 与 `comet\u002Feval.yaml`。\n8. 验证：展示 quick\u002Ffull eval 工作量，运行或记录当前 draft hash 的 eval evidence；失败、skip 或证据 hash 过期时不得进入 ready。\n9. Review \u002F readiness：读取 `comet publish review \u003Cname> --platform \u003Creference-platform> --json`，展示 `Readiness:`、`Blockers:`、`Warnings:`、`Evidence:`。\n10. Publish \u002F install preview：人工批准后才能 publish；安装前必须先运行 preview，并展示 `No files were written`。\n\n## 方案示例\n\n组件库和白盒审查场景应生成类似 plan：\n\n```json\n{\n  \"goal\": \"基于 \u002Fcomet-classic 的五阶段定制，要求组件库和白盒审查。\",\n  \"skillCreatorIntent\": \"customize-comet\",\n  \"workflow\": {\n    \"kind\": \"comet-five-phase-overlay\",\n    \"name\": \"team-comet\",\n    \"goal\": \"要求组件库和白盒审查。\",\n    \"nodes\": {\n      \"execute\": {\n        \"requiredSkillCalls\": [\n          {\n            \"skill\": \"elementui\",\n            \"reason\": \"Use project component library during direct implementation.\"\n          }\n        ]\n      },\n      \"subagent-execute\": {\n        \"requiredSkillCalls\": [\n          {\n            \"skill\": \"elementui\",\n            \"scope\": \"handoff\"\n          }\n        ]\n      },\n      \"review\": {\n        \"requiredSkillCalls\": [\n          {\n            \"skill\": \"whitebox-code-standard\",\n            \"scope\": \"review\"\n          }\n        ]\n      }\n    }\n  }\n}\n```\n\n## 硬性规则\n\n- 必须先展示方案确认页，再生成。\n- 确认页必须为每个新增 binding 或 schema 显示 enforcement：`guarded`、`handoff-guarded`、`evidence-only` 或 `advisory`。\n- Required Skill Call 不替换 Node implementation。\n- producer override 必须声明 `satisfies` 的 Output Schema。\n- Output Schema 必须挂到具体 Workflow Node 才算生效；只定义在 `workflow.outputSchemas` 里不会触发 guard、eval 或 readiness。\n- control Node 普通模式不得 override。\n- eval、review、publish readiness 必须读取同一份 `workflow-protocol.json`。\n- readiness blockers 必须阻止 publish：缺少当前 draft hash 的 eval evidence、人工 approval、required capability 或 executable disclosure 任一项都不能进入 ready。\n- 子代理 Handoff 必须要求子代理加载 Required Skill Call 并回传 evidence。\n- 脚本只读取 protocol 和 state，不把 Skill 名称当成校验依据。\n- 安装前必须询问用户，不得自动安装。\n\n## 参考资料\n\n- `comet-any\u002Freference\u002Fauthoring-protocol.json`\n- `comet-any\u002Freference\u002Fauthored-zone-example.md`\n- `comet-any\u002Freference\u002Fbundle-authoring.md`\n- `comet-any\u002Freference\u002Fauthoring-subagents.md`\n- `comet-any\u002Freference\u002Feval-provider.md`\n",{"name":5,"description":7,"disable-model-invocation":24},true,"master","480007470440d896436e78c609ab33fcd24ec73f",".agents\u002Fskills\u002Fcomet-any",8,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime","https:\u002F\u002Fdeepyou.top","https:\u002F\u002Fgithub.com\u002Fcocoyes","You are not too good, I like it .","2026-11-01 17:25:06",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":24},[36,39,42],{"tool":37,"command":38},"npx","npx skills add cocoyes\u002Fzhizhi-agent-runtime --skill comet-any",{"tool":40,"command":41},"bunx","bunx skills add cocoyes\u002Fzhizhi-agent-runtime --skill comet-any",{"tool":43,"command":44},"pnpm","pnpm dlx skills add cocoyes\u002Fzhizhi-agent-runtime --skill comet-any","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002FSKILL.md",[47,53,57,61,65,69,73,77,81,85,89],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"agents\u002Fopenai.yaml","REFERENCE","参考文档",166,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Fagents\u002Fopenai.yaml",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"reference\u002Fauthored-zone-example.md",8362,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fauthored-zone-example.md",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"reference\u002Fauthoring-subagents.md",5266,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fauthoring-subagents.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"reference\u002Fbundle-authoring.md",6463,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fbundle-authoring.md",{"path":66,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":67,"sourceHtmlUrl":68},"reference\u002Feval-provider.md",3218,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Feval-provider.md",{"path":70,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":71,"sourceHtmlUrl":72},"reference\u002Fsubagents\u002Fpause-points-author.md",4313,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fsubagents\u002Fpause-points-author.md",{"path":74,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":75,"sourceHtmlUrl":76},"reference\u002Fsubagents\u002Freference-author.md",3071,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fsubagents\u002Freference-author.md",{"path":78,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":79,"sourceHtmlUrl":80},"reference\u002Fsubagents\u002Fscript-author.md",3614,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fsubagents\u002Fscript-author.md",{"path":82,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":83,"sourceHtmlUrl":84},"reference\u002Fsubagents\u002Fskill-core-author.md",7100,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fsubagents\u002Fskill-core-author.md",{"path":86,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":87,"sourceHtmlUrl":88},"reference\u002Fsubagents\u002Fskill-reviewer.md",6575,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fsubagents\u002Fskill-reviewer.md",{"path":90,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":91,"sourceHtmlUrl":92},"reference\u002Fsubagents\u002Fworkflow-entry-author.md",6225,"https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-any\u002Freference\u002Fsubagents\u002Fworkflow-entry-author.md",[94,101,108,115,122],{"skillNo":95,"name":96,"nameZh":6,"description":97,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":98,"categories":99,"keywords":100,"publishedAt":19},"01M7BV5DE31NVJ8RMSJEZ81DCR","test-skill","Execute the audited Python hello script and create a small report file.","https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002Fexamples\u002F15-docker-skills\u002Fskills\u002Ftest-skill\u002FSKILL.md",[],[],{"skillNo":102,"name":103,"nameZh":6,"description":104,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":105,"categories":106,"keywords":107,"publishedAt":19},"01M7BV5C8KAA49KBYDR0QGQXYH","greeting","Create a greeting using the application greeting tool.","https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002Fexamples\u002F14-skills\u002Fskills\u002Fgreeting\u002FSKILL.md",[],[],{"skillNo":109,"name":110,"nameZh":6,"description":111,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":114,"publishedAt":19},"01M7BV5AS8ZWVNQBB80VRE6XRY","comet","Comet 工作流入口。当用户明确调用 \u002Fcomet，或明确要求使用 Comet 但未指定 Native\u002FClassic 时使用；按项目配置加载 Native 或 Classic。","https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet\u002FSKILL.md",[],[],{"skillNo":116,"name":117,"nameZh":6,"description":118,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":119,"categories":120,"keywords":121,"publishedAt":19},"01M7BV597B559GP5MDDSW4PAYM","comet-review","手动审查当前 Comet change 的实现差异，只报告正确性、安全和边界问题，不推进工作流。","https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-review\u002FSKILL.md",[],[],{"skillNo":123,"name":124,"nameZh":6,"description":125,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":126,"categories":127,"keywords":128,"publishedAt":19},"01M7BV57E4RXGBJEP7J8NBH8WA","comet-native","Comet Native 工作流。当用户明确调用 \u002Fcomet-native、要求启动或恢复 Native change，或入口路由到 Native 时使用。","https:\u002F\u002Fgithub.com\u002Fcocoyes\u002Fzhizhi-agent-runtime\u002Fblob\u002Fmaster\u002F.agents\u002Fskills\u002Fcomet-native\u002FSKILL.md",[],[]]