一键导入
flow-code-guide
Skill discovery guide. Use at session start or when unsure which skill to apply. Shows the flowchart for selecting the right flow-code skill.
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。
菜单
Skill discovery guide. Use at session start or when unsure which skill to apply. Shows the flowchart for selecting the right flow-code skill.
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
Internal pipeline engine. Manages the entire pipeline (brainstorm, plan, plan-review, work, impl-review, close) via flowctl phase commands. Invoked by /flow-code:go.
Use when exploring requirements before planning. Pressure-tests ideas, generates approaches, and outputs a requirements doc for /flow-code:plan. Supports --auto mode for AI self-interview (no human input needed).
Use when exploring requirements before planning. Pressure-tests ideas, generates approaches, and outputs a requirements doc for /flow-code:plan.
Use when planning features or designing implementation. Triggers on /flow-code:plan with text descriptions or Flow IDs.
Use when reviewing code changes — self-review in Worker Phase 6, impl-review, or PR review. Applies five-axis scoring with severity labels.
Use when implementing a plan or working through a spec. Triggers on /flow-code:work with Flow IDs.
| name | flow-code-guide |
| description | Skill discovery guide. Use at session start or when unsure which skill to apply. Shows the flowchart for selecting the right flow-code skill. |
| tier | 1 |
| user-invocable | true |
Full autopilot (idea to PR): /flow-code:go "your idea"
Individual phases:
/flow-code:brainstorm "idea"/flow-code:go "idea" --plan-only/flow-code:go fn-1-add-oauthWhat are you trying to do?
│
├─ Build a feature from scratch?
│ └─ /flow-code:go "description"
│ (runs: brainstorm → plan → work → review → close)
│
├─ Plan without implementing?
│ └─ /flow-code:go "description" --plan-only
│
├─ Resume existing work?
│ └─ /flow-code:go fn-N-slug
│
├─ Explore an idea before committing?
│ └─ /flow-code:brainstorm "idea"
│
├─ Build/modify UI components?
│ └─ /flow-code:frontend-ui "component description"
│ (design system, accessibility, AI aesthetic avoidance)
│
├─ Simplify complex code?
│ └─ /flow-code:simplify "file or directory"
│ (Chesterton's Fence, behavior preservation)
│
├─ Debug a bug or test failure?
│ └─ Use flow-code-debug methodology
│ (reproduce → localize → reduce → fix → guard)
│
├─ Audit codebase readiness?
│ └─ /flow-code:prime
│
├─ Map codebase architecture?
│ └─ /flow-code:map
│
├─ QA test a web app?
│ └─ /flow-code:qa "url"
│
├─ Visual design audit?
│ └─ /flow-code:design-review "url"
│
└─ Improve code autonomously?
└─ /flow-code:auto-improve "goal" --scope dir/
These skills are loaded during /flow-code:go based on task domain. You don't invoke them directly — they guide worker behavior:
| Skill | Loaded When | Core Methodology |
|---|---|---|
| Core (all tasks) | ||
flow-code-incremental | All tasks | Vertical slicing, Implement→Test→Verify→Commit |
flow-code-code-review | All tasks (Phase 6) | Five-axis review, severity labels |
flow-code-error-handling | All tasks with error paths | Retry, circuit breaker, graceful degradation |
| Frontend | ||
flow-code-frontend-ui | domain=frontend | Component architecture, WCAG 2.1 AA |
flow-code-state-management | domain=frontend | State classification, decision ladder, server state |
flow-code-i18n | i18n/l10n tasks | ICU messages, Intl APIs, RTL support |
| Backend | ||
flow-code-api-design | domain=backend/architecture | Contract-first, Hyrum's Law, error semantics |
flow-code-security | domain=backend/architecture/ops | OWASP Top 10, three-tier boundaries |
flow-code-database | DB/ORM tasks | Migration safety, N+1 detection, indexing |
flow-code-auth | Auth tasks | OAuth, JWT lifecycle, RBAC, resource-level auth |
flow-code-caching | Caching tasks | Cache-aside, TTL, invalidation strategies |
| Infrastructure | ||
flow-code-observability | domain=backend/ops | Structured logging, tracing, metrics (RED/USE) |
flow-code-monitoring | domain=ops | SLO/SLI, dashboards, alerting, runbooks |
flow-code-containerization | Docker/K8s tasks | Multi-stage builds, health probes, security |
flow-code-microservices | Service boundary tasks | Saga, events, data ownership, decomposition |
| Quality | ||
flow-code-tdd | --tdd flag or domain=testing | Red-Green-Refactor, Prove-It Pattern |
flow-code-debug | domain=testing or bug tasks | Reproduce→Localize→Reduce→Fix→Guard |
flow-code-performance | Performance tasks | Measure→Identify→Fix→Verify→Guard |
flow-code-simplify | /flow-code:simplify or auto-improve | Chesterton's Fence, 18 patterns |
flow-code-documentation | docs domain or releases | ADRs, README, API docs, changelogs |
Quick-lookup docs for reviews and verification:
/flow-code:go "idea"
├── Brainstorm (AI self-interview)
├── Plan (scouts + task creation)
├── Plan Review (RP/Codex)
├── Work (parallel workers in worktrees)
│ └── Worker phases: investigate → implement → verify → commit → review
├── Impl Review (adversarial)
└── Close (pre-launch checklist → PR)