跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
设计与 UI

recipe-front-design

Execute from codebase-scoped analysis to frontend design document creation including UI Spec.

shinpr · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add shinpr/codex-workflows --skill recipe-front-design

技能说明


name: recipe-front-design description: "Execute from codebase-scoped analysis to frontend design document creation including UI Spec."

Context: Dedicated to the frontend design phase.

Required Skills [LOAD BEFORE EXECUTION]

  1. [LOAD IF NOT ACTIVE] documentation-criteria -- document quality standards
  2. [LOAD IF NOT ACTIVE] implementation-approach -- design convergence and verification strategy
  3. [LOAD IF NOT ACTIVE] subagents-orchestration-guide -- agent coordination and review resolution
  4. [LOAD IF NOT ACTIVE] llm-friendly-context -- document and review handoffs

Load external-resource-context in Step 4 only when a named external source is required for the current design or verification decision.

Spawn rule: every spawn_agent call uses fork_turns="none" so the subagent receives only the task message and explicitly provided context.

Orchestrator Definition

Core Identity: Coordinate frontend design, make workflow decisions from compact specialist materials, and invoke specialists for analysis, authoring, and review.

Execution Plan: Reuse the active execution plan. When the workflow has multiple dependent actions and no plan exists, create one that tracks them through final verification.

Execution Method:

  • Scope and cost evidence -> performed by requirement-analyzer
  • Codebase analysis -> performed by codebase-analyzer
  • Scope confirmation -> performed by the orchestrator with user confirmation
  • UI fact gathering -> performed by ui-analyzer
  • UI Specification creation -> performed by ui-spec-designer
  • Design document creation -> performed by technical-designer-frontend
  • Design Doc verification -> performed by code-verifier
  • Document review -> performed by document-reviewer

Orchestrator spawns agents and passes structured data between them.

Scope Boundaries

Included in this skill:

  • Compact scope and cost evidence from requirement-analyzer; the orchestrator owns requirement, scale, UI scope, and ADR decisions
  • Codebase analysis with codebase-analyzer (entry point of the frontend design phase)
  • Scope confirmation with the user, grounded in codebase-analyzer findings
  • Focused external resource hearing when a current design decision requires it
  • UI fact gathering with ui-analyzer
  • UI Specification creation with ui-spec-designer (prototype code inquiry included)
  • One ADR per qualifying decision point found in the current scope, created and reviewed as one batch
  • Design Doc creation with technical-designer-frontend
  • Document review with document-reviewer

Responsibility Boundary: This skill completes with approval of the UI Spec, Design Doc, and its preceding ADR when required. Work planning and beyond are outside scope.

Requirements: $ARGUMENTS

Execution Flow

Step 1: Scope and Cost Evidence

Spawn requirement-analyzer with the original requirements. Treat its request signals, scope evidence, cost evidence, and questions as material; the orchestrator determines requirements, scale, UI scope, and ADR routing.

Step 2: Codebase Analysis

Spawn codebase-analyzer agent: "exploration_mode: [mode from Analysis Assignment]. Analyze the existing codebase to provide compact decision materials for frontend requirement confirmation, ADR selection, minimal Design Doc creation, and verification. requirement_analysis: [Step 1 scopeEvidence]. requirements: [original user requirements]. layer: frontend. target_paths: [Step 1 scopeEvidence.affectedFiles]. focus_areas: responsibility ownership, state/data paths, contracts, and reuse."

Step 3: Scope Confirmation

After codebase-analyzer returns, confirm requirements, determine Structural Scale, and collect candidate ADR decision points:

  1. Locate a related PRD and read its Converged Outcome, MVP scope, Future / Out of Scope, and open requirement fields. If the related PRD is ambiguous, ask the user to select or provide its path, or confirm none exists, before continuing.
  2. When those fields match the current request and returned scope facts, use the PRD path as the current carrier and proceed directly to scope confirmation.
  3. When a current carrier is absent, load requirement-convergence. The orchestrator builds and judges its record from the user's wording, using Step 1 scope/cost evidence and Step 2 analysis for trade-offs, questions, and routing decisions. Mark an existing but incomplete or scope-mismatched PRD for update; otherwise mark the carrier as absent.
  4. Retain Step 2 decisionMaterials.candidateDecisionPoints within confirmed scope. Final ADR qualification occurs after UI analysis and UI Spec evidence are available.
  5. Determine Structural Scale and set prdRequired when the scale is Large and the current PRD carrier is absent.

Present the frontend design scope to the user:

  • Target files/modules: analysisScope.filesAnalyzed and directly relevant components, routes, or modules
  • Affected layers: analysisScope.affectedLayers
  • Recommended document path: UI Spec and Design Doc, plus a conditional ADR batch for listed candidate decision points that later pass both ADR filters
  • PRD status: whether prdRequired is true and whether the convergence carrier is current, requires update, or is absent
  • Unknowns/assumptions: limitations and unresolved risks
  • Questions before design: scope questions that change the UI surface, design target, or scale, including technical wording whose mandatory/candidate status is outcome-relevant and ambiguous

Ask the user to choose one:

  • Proceed with the recommended document path
  • Correct the scope and re-run codebase-analyzer
  • Answer open questions, then proceed
  • Provide an existing PRD path when prdRequired is true
  • Explicitly approve proceeding without a PRD when prdRequired is true and no PRD will be provided

If prdRequired is true and the user neither provides a PRD path nor explicitly approves proceeding without a PRD, stop. This recipe does not create PRDs.

[STOP -- BLOCKING] Wait for us

阅读完整来源 ↗