菜单
Pre-task token discipline. Use before any task when every read, tool call, line of reasoning, and output token must materially advance the stated goal or thesis, without weakening correctness or safety.
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
Brief pre-work alignment for any stage. Use before planning, coding, editing, reviewing, or delegating when the user asks you to understand the thesis, vibe, intent, current position, or direction of a request or existing thing. Returns 2–4 bold-labelled one-sentence observations and one short synthesis, then stops.
The single front door to the SDD pipeline (docs/plans/): research → plan → workshop → ADR → phase tasks → implement → progress → review → ship. Use when the user wants to plan, research, explore, specify or architect (now one atomic 'plan' step → one document: business spec + implementation plan), clarify, workshop, write an ADR, break a phase into tasks, implement or build a phase (optionally with a live review companion), update progress, code-review, or ship a plan (push + PR + watch checks) — or types /the-flow (or a legacy /plan-N command), or asks to start, resume, or adopt a plan flow. Guided mode (no args) coaches from durable on-disk state; direct jump runs one stage: /the-flow <id> <verb> [flags] — 1a explore, 1b plan (spec + plan in one doc; incl. clarify), 2c workshop, 3a adr, 5 tasks, 6 implement (--companion for live review), 6a progress, 7 review, 8 ship, 8c reconcile. Ids and verbs resolve alone; printed commands carry both.
Adaptive, high-signal post-action validation for whatever was just produced — unified plan, tasks, code change, workshop, legacy spec, or any artifact. Proves the claims that matter, reports only findings likely to change action, and adds independent critics only when risk or uncertainty earns them. Thesis- and consumer-aware; read-only through adjudication; compact verdict.
FlowSpace-first codebase research agent for parallel subagent exploration with structured output for parent orchestrator synthesis. Domain-aware v2.
Establish or refresh the project constitution and align the supporting norms documents before any planning phases begin. Re-entrant command that preserves user customizations during updates. Domain-aware v2.
Generate terse, industry-standard issue text from specs and plans for Azure DevOps, GitHub Issues, or any tracker. Domain-aware v2.
| name | frugal |
| description | Pre-task token discipline. Use before any task when every read, tool call, line of reasoning, and output token must materially advance the stated goal or thesis, without weakening correctness or safety. |
Use as a silent pre-task constraint. Keep the stated goal or thesis in front; before any file read, search, tool call, line of code, note, or explanation, ask whether it can change a decision, constrain implementation, prove behavior, reduce risk, preserve intent, or enable the next action. Skip broad context, decorative prose, duplicate summaries, speculative branches, and "while I'm here" work; read narrowly, expand only when uncertainty is material, and prefer direct evidence, existing primitives, and the smallest check or artifact that advances the goal. Do not economize away correctness, security, data-loss prevention, accessibility, user intent, or irreversible-action warnings. No ceremony: apply this silently and output only the next useful thing, then stop.
