菜单
Use when planning, building, or reviewing a software milestone/release/version — any task framed as "ship vX", "build the next milestone", "plan and execute a feature set", "do a research + build pass", or "review what we shipped".
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when orchestrating multi-agent or long-running work — spawning sub-agents, fanning out parallel tasks, delegating subtasks, running build-then-review loops, or deciding which model tier should do a piece of work.
Use when building a self-contained shippable artifact from scratch or a major rev — an MCP server, a CLI, a library/SDK, a skills pack, a plugin — anything that gets packaged and released rather than merged into an existing app.
Use when maintaining project state across sessions — writing or reading a STATE.md / MEMORY.md, recording what was learned, resuming interrupted work, or deciding what to persist for future sessions.
Use when building or redesigning a product UI/visual system — creating a design system, writing or following a DESIGN.md contract, generating screens in a design tool (Stitch/Figma) then porting to code, or doing a branded visual redesign.
Use during hands-on execution of any coding or agent task — especially when following multi-item instructions, deciding how much to deliberate, about to claim something is done, or producing output at risk of reading as generic AI slop.
Use when building or running an automated measurement/audit/reporting pipeline that scores many subjects on a recurring basis and reports to stakeholders — a compliance monitor, a benchmark matrix, a metrics dashboard, a scheduled scan-and-report.
| name | milestone-workflow |
| description | Use when planning, building, or reviewing a software milestone/release/version — any task framed as "ship vX", "build the next milestone", "plan and execute a feature set", "do a research + build pass", or "review what we shipped". |
A disciplined way to take a milestone from idea to shipped, distilled from real high-quality milestone cycles. Three phases, each with a dedicated command:
/milestone-research) — figure out what's worth
building before writing code; produce an approved design doc./milestone-build) — execute the design in atomic, verified steps./milestone-review) — a cross-perspective pass that
catches what self-review can't, before declaring done.Run them in order across a milestone. You don't have to use the commands — when you recognize a phase, apply its discipline directly.
This workflow is generic. Before acting, read the repo's CLAUDE.md (and
.planning/STATE.md / ROADMAP.md if present) to learn the project-specific:
Never hardcode another project's commands. If CLAUDE.md doesn't state a verify
command, ask or infer from package.json scripts — don't guess silently.
These are the through-lines that made the reference cycles work:
The full, paste-ready instructions for each phase live in references/:
references/phase-research.mdreferences/phase-build.mdreferences/phase-review.mdThe /milestone-* commands are thin entry points to these same files, so the
skill and commands never duplicate the prompt text.
This skill is the generalized form of the patterns observed across real milestone cycles (research-rounds + cross-AI consult → P0 design → phased build with verify-after-each-wave → numbered adversarial-fix batches). It is the HOW; each repo's CLAUDE.md + design doc is the WHAT.