ワンクリックで
workflow-bootstrap
Use when defining a tool-agnostic workflow shell, role split, and future runtime-pack mapping for a repository.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Use when defining a tool-agnostic workflow shell, role split, and future runtime-pack mapping for a repository.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Adapter entry for bootstrapping agent rules, contracts, validation, task packages, reports, and handoff boundaries for financial data engineering or operations projects. Canonical source is in skills/financial-data-agent-bootstrap.
Use when bootstrapping agent rules, contracts, validation, task packages, reports, and handoff boundaries for financial data engineering or operations projects.
Use when evaluating, scoring, diagnosing, or explicitly refactoring a single skill under controlled governance.
Use when onboarding a new maintainer or AI agent, scanning an unfamiliar repository, or generating a takeover packet.
Use when preparing a handoff task package, bounded execution, or execution report across AI agents.
Use when assessing migration readiness for a financial-data Python project with script sprawl, Excel assets, or Wind/Desktop coupling.
| name | workflow-bootstrap |
| description | Use when defining a tool-agnostic workflow shell, role split, and future runtime-pack mapping for a repository. |
| metadata | {"triggers":["bootstrap a role-based workflow","define drafter reviewer implementer reporter final reviewer boundaries","map canonical guidance to a future runtime pack","explain thin entry versus full canonical guidance","document workflow shell boundaries without replacing handoff protocols"],"side_effects":["read_only","write_files"]} |
This execution-focused skill definition keeps the workflow-shell behavior, invocation shape, and adapter-facing contract in one canonical location while moving explanation-oriented detail into supporting assets.
这个 pattern 的核心是先把默认协作链路和角色边界说清,再把后续执行交给更专门的协议或治理 skill。
Drafter -> Reviewer -> Implementer -> Reporter -> Final ReviewerInput:
Process:
chatgpt-handoff-pilotOutput:
chatgpt-handoff-pilot 的 protocol ownership这样组织的原因是,很多多 agent 协作问题并不是缺一个新工具,而是缺一个清晰的默认链路。workflow-bootstrap 负责补齐这层壳,但不接管已有协议层、接管层、状态层或治理层的职责。
skills/ 始终是唯一 canonical source of truth。
Keep skills/ as the only canonical source.
默认使用工具无关的 role chain;具体工具只作为可替换 adapter examples。 Use the role chain as canonical guidance; treat concrete tools as replaceable adapter examples.
Drafter、Reviewer、Implementer、Reporter、Final Reviewer 只做职责拆分,不额外发明控制层。 Keep role split minimal instead of introducing a new control framework.
handoff 协议继续复用 chatgpt-handoff-pilot。
Reuse chatgpt-handoff-pilot for task packages, bounded execution, and execution reports.
在 non-git / low-git 项目侧试跑里,tasks/ 可作为 task package / execution report 的主证据目录。
In non-git / low-git project-side trials, tasks/ may serve as the primary trace path for task packages and execution reports.
docs/HANDOFF.md、status surface 与 archive/ 只承接最小闭环或历史参考,不应取代 active workflow line。
Keep docs/HANDOFF.md, status surfaces, and archive/ limited to minimal closure or historical reference rather than the active workflow line.
薄入口只负责发现与回指,不复制 canonical 细节。
Use thin entries for discoverability and redirection, not full duplication.
runtime pack guidance 只能写成项目侧 thin surface guidance,不能写成当前 hub 已落地事实或第二规则库。 Describe runtime-pack assets as project-side thin surface guidance only.
判断当前需求是否属于 workflow bootstrap。
当目标是定义默认链路、角色分工、runtime-pack 映射,或为后续项目侧入口做 canonical 对齐时,优先使用本 skill;如果目标其实是 handoff、takeover、status update、documentation governance 或 structure audit,则转向对应 skill。
对齐默认链路。
明确默认路径为 Drafter -> Reviewer -> Implementer -> Reporter -> Final Reviewer。具体工具名只作为 adapter examples;role-based handoff 详见 role_split_and_integration.md。
明确最小职责边界。
Drafter 负责收敛方案与任务边界;Reviewer 负责 implementation 前的 safety gate;Implementer 负责按任务包做 bounded execution;Reporter 负责 execution report;Final Reviewer 负责边界、验证与闭环检查。不要让本 skill 取代具体实现技能或治理技能。
衔接 handoff 协议。
当实施工作需要 task package、bounded execution 和 execution report 时,显式复用 chatgpt-handoff-pilot。本 skill 只定义 workflow 壳层,不重新定义 handoff 协议。
对齐 non-git / low-git 证据约定。
当项目侧试跑缺少稳定 Git 证据时,可把 tasks/ 作为 task package / execution report 的主 trace path,并把 execution report 作为每轮实施的主证据。若 workflow 输出需要 assessment / status / handoff 衔接意识,可引用 shared assessment output protocol 的字段口径;docs/HANDOFF.md 与 status surface 只保留最小闭环事实,archive/ 只作为历史参考,不恢复为 active workflow line。
映射 runtime pack guidance。
当需要为项目侧协作入口做准备时,只描述 AGENTS.md、.github/copilot-instructions.md、.github/copilot-instructions.zh-CN.md、tasks/README.md、task package 和 execution report 等最小候选 surface 与 canonical guidance 的关系,不在 hub 内创建这些项目侧文件。
保持 discoverability 薄层。
若新增或维护 adapter 入口,应保持 .agents/skills/ 与 .github/skills/ 为薄封装或兼容入口,并回指 canonical skills/workflow-bootstrap/,不要复制大段 canonical 内容。
输出简洁实施结果。
最终结果应说明:workflow 壳层是否已对齐、与现有相关 skills 的边界是否清楚、discoverability 是否补齐、以及哪些内容被明确保留到 future runtime pack 阶段。
chatgpt-handoff-pilot 的 task package / bounded execution / execution report 协议project-takeover 的接管与 onboarding 输出update-project-status 的状态刷新职责documentation-governance 的文档治理职责file-structure-check 的结构审计职责docs/HANDOFF.md、status surface 或 archive/ 写成 per-task 主证据线.agents/skills/ 或 .github/skills/ 升级为第二事实源