菜单
Specification-workflow patterns (input assessment, produce-then-validate, gap-informed revision) consumed by the State Analyst alongside strategy-core for targeted Supervisor briefings.
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
| name | strategy-specification |
| description | Specification-workflow patterns (input assessment, produce-then-validate, gap-informed revision) consumed by the State Analyst alongside strategy-core for targeted Supervisor briefings. |
Workflow-specific patterns for the specification workflow. Consumed by the State Analyst alongside strategy-core to produce targeted Supervisor briefings.
Produce a validated feature specification where the advocate verdict is ready.
spec.md exists with user stories and functional requirementsadvocate-report.md verdict: readySparse input (missing Who/Problem/Value) typically benefits from enrichment before analysis. Detailed input with clear problem framing can go directly to the analyst. Consider the domain context — a sparse input in a well-understood domain may not need enrichment if the analyst can infer context from existing artifacts (constitution, codebase analysis).
Rationale: Enrichment adds a full interaction round with the user. When the input already contains sufficient signal, skipping enrichment saves time without sacrificing quality. The analyst is capable of working with moderately detailed input.
The primary pattern is: an agent produces an artifact, then a different agent (or gate) validates it. For specification, this means the analyst produces spec.md, the advocate reviews it. Resist the urge to skip validation even when the analyst report looks comprehensive.
Rationale: Self-review is unreliable. The analyst's blind spots are systematic — a separate reviewer with adversarial framing catches what the producer cannot see in their own work.
When the advocate identifies gaps, the next analyst pass should be informed by those specific gaps. The analyst should see what was flagged and focus revision effort there, not rewrite the entire spec. Use informed-by edges to carry gap context forward.
Rationale: Unfocused revision wastes effort on sections that were already adequate, and may introduce new issues in previously-clean areas. Targeted revision converges faster.
When gaps are knowledge-based (factual unknowns about the codebase, existing protocols, technical constraints), targeted research often resolves them faster than asking the user. Try research first; escalate to user only if research is inconclusive or the question is genuinely about preference.
Rationale: Users should not be asked questions that the codebase can answer. Research is faster and more precise for factual questions. Reserve user interaction for genuine decisions.
This skill MUST be invoked when the user says "implementation strategy", "cycle sequencing", or "implement workflow patterns". SHOULD also invoke when user mentions "execute-then-verify", "targeted retry", "fix pass", or "implementation escalation". Provides implementation-workflow patterns consumed alongside strategy-core for targeted briefings.
This skill MUST be invoked when the user says "audit agent", "review agent", "check agent compliance", or "agent audit checklist". SHOULD also invoke when user mentions "AGENT-GUIDELINES compliance", "agent review", or "coupling detection". Provides the procedural framework for auditing agents against AGENT-GUIDELINES.md.
This skill MUST be invoked when the user says "execute cycle", "implement tasks", "TDD cycle", or "red green refactor". SHOULD also invoke when implementing cycle task lists, encountering `[EXTEND]`/`[MODIFY]` markers, generating cycle reports, or handling retry after checkpoint failure.
This skill MUST be invoked when the user says "create task mapping", "structure implementation", "define cycles", "vertical slice", "TDD", "test first", "cycle structure", or "testable increment". SHOULD also invoke when user mentions "red green refactor" or "implementation tasks".
This skill MUST be invoked when the user says "TEST:", "TEST:VERIFY", "TEST:CONTRACT", "execute verification", "run test task", or "verification test". SHOULD also invoke when the user says "Setup/Action/Assert", "verify against infrastructure", or "capture evidence".
This skill MUST be invoked when the user says "write technical requirements", "define constraints", "define NFRs", "map integrations", "classify data sensitivity", or "infrastructure requirements". SHOULD also invoke when user mentions "TR-", "C-", "NFR-", "IP-", "INT-", "DS-", "non-functional", "system integration", "infrastructure provisioning", or "data classification". Produces three traceable analysis artifacts from business specifications.