菜单
FOUNDATIONAL SKILL - Analyze every prompt for underlying motivation using Golden Circle (WHY->HOW->WHAT). Runs BEFORE any other skill. No action without understanding WHY.
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。
AID Phase 4 - Development phase. Use for implementing features, TDD practices, code reviews, transitioning from planning to QA.
AID Phase 0 - Research & discovery. Use for validating problem spaces, identifying stakeholders, defining success metrics, deciding whether to proceed.
AID Phase 3 - Implementation Planning with consolidation-first approach. Resolves contradictions between PRD and Tech Spec, creates consolidated master document, then breaks down into actionable tasks and populates Jira. Includes sprint planning and risk assessment.
AID Phase 1 - PRD creation. Use for user stories, acceptance criteria, scoping features, transitioning from discovery to tech spec.
AID Phase 5 - QA and Release. Use for validating implementations, acceptance tests, preparing releases, deployment, operational readiness.
AID Phase 2 - Technical Specification. Use for system architecture, API contracts, data models, security architecture, transitioning from PRD to implementation.
| name | why-driven-decision |
| description | FOUNDATIONAL SKILL - Analyze every prompt for underlying motivation using Golden Circle (WHY->HOW->WHAT). Runs BEFORE any other skill. No action without understanding WHY. |
EVERY PROMPT -> WHY ANALYSIS -> THEN PROCEED
Never act without understanding WHY.
WHY <- Purpose, belief, motivation
HOW <- Process, values, approach
WHAT <- Features, outputs, deliverables
Work inside-out: WHY -> HOW -> WHAT
1. EXTRACT EXPLICIT WHY - What did user state as goal?
2. INFER IMPLICIT WHY - What need not articulated?
3. VALIDATE OR ASK - Is WHY clear enough?
4. ANCHOR TO WHY - Every output traces to WHY
PROMPT: "Add loading spinner to dashboard"
STEP 1: Explicit WHY: Not stated
STEP 2: Implicit WHY: Users confused? App frozen? Duplicate clicks?
STEP 3: ASK: "What problem? Users think broken, or clicking multiple times?"
STEP 4: Anchor all decisions to answer
| Question | Answer |
|---|---|
| WHY am I doing this? | Purpose |
| WHAT value created? | Benefit |
| WHO benefits? | Stakeholder |
Can't answer in 3 seconds -> STOP and clarify.
"We need a dashboard"
Why? -> "To see metrics"
Why? -> "To track performance"
Why? -> "To make better decisions"
Why? -> "Because overwhelmed by data"
ROOT: Reduce overwhelm, create clarity
| Phase | Core Question |
|---|---|
| Discovery | "Why is this problem worth solving?" |
| PRD | "Why does user need this feature?" |
| Tech Spec | "Why this architecture?" |
| Development | "Why this code? Why these connections?" |
| QA & Ship | "Why this test? Why ready?" |
| Signal | Ask |
|---|---|
| "Just do it" | "What problem does this solve?" |
| "Everyone wants it" | "Why specifically?" |
| "Competitor has it" | "Does our WHY require it?" |
| "It's urgent" | "What's cost of not doing it?" |
| "Trust me" | "Help me understand reasoning" |
# WHY: Users with slow connections timing out on large datasets.
# WHAT: Paginated query with streaming response.
# CONNECTION: Called by ReportGenerator, feeds into ExportService.
def paginated_query(query: str, page_size: int = 100) -> Iterator[Row]:
"""
WHY page_size=100: Balance memory (<10MB) vs latency (<50ms).
WHY Iterator: Allows early stop without loading all data.
"""
def test_retry_on_network_timeout(self):
"""
WHY: 3% of payments fail due to network issues.
EXPECTED: Retry 3x recovers 90%.
COST OF FAILURE: $15 avg transaction lost.
"""