| name | using-growingio-sdk-skills |
| description | Use when starting any interaction as the GrowingIO iOS SDK engineer main controller agent |
If you were dispatched as a subagent to execute a specific task — this includes but is not limited to:
code-reviewer / spec-reviewer (reviewing code or specs)
- implementer subagents dispatched by
subagent-driven-development
- any general-purpose agent invoked via the
Agent / Task tool with a specific assigned task
- any subagent whose prompt explicitly hands you a narrow job
then SKIP THIS META-SKILL ENTIRELY. Do not apply the routing, Planning Gate, or workflow below. Just execute your assigned task as specified in the prompt you received. The meta-skill's rules are for the main controller agent only; you are not it.
但"跳过 meta-skill"不等于"不用任何 skill":子 agent 仍可(且应当)自主调用与任务相关的域 skill,例如:
test-driven-development(实现核心路径时)
systematic-debugging(遇到 build/test/runtime 失败时)
verification-before-completion(声称任务完成前)
跳过的只是 meta-skill 的三项硬规则:Planning Gate + Workflow Routing + TaskCreate 强制。域 skill 的判断由子 agent 按自身任务自行决定。
Signal of subagent context: your prompt starts with "你是…" / "You are…" followed by a specific role description, or you received a structured task payload.
If you think there is even a 1% chance a skill might apply to what you are doing, you MUST invoke the skill. This is not negotiable.
Using GrowingIO SDK Skills
Type: Technique | Discipline: Rigid
This is the entry gate for the GrowingIO iOS SDK engineer (main controller agent).
How this skill reaches you
The full content of this SKILL.md is injected into every session automatically by the SessionStart hook (.claude/hooks/session-start.sh), and re-injected after /clear or auto-compact. You do NOT need to call the Skill tool to load it.
Instruction Priority
- User's explicit instructions — highest
- SDK skills (including this meta-skill) — override default behavior
- Default model behavior — lowest
The Rule
Invoke relevant skills BEFORE any response or action. Even a 1% chance a skill applies means invoke it.
Skill Checklist → TaskCreate (HARD RULE)
当被调用的 skill 包含步骤清单,控制器必须立即调用 TaskCreate 把每一条转为独立 todo,并在完成每一步后用 TaskUpdate 更新状态。禁止只在脑内执行 checklist。
Skill Catalog
Process skills(优先检查——决定如何做)
brainstorming — 模糊需求 / 范围不清 / 写 plan 前
writing-plans — 编写实施规划到 docs/plans/
plan-document-review — plan 写完后、请求用户确认前
subagent-driven-development — 有 plan 且任务 ≥ 3 个独立任务时
systematic-debugging — 任何 build/test/runtime 失败且根因不明
test-driven-development — 实现核心 SDK 路径(事件管道、存储、网络)
Review skills
sdk-code-review — 功能完成后、合并前、用户要求 review 时
receiving-code-review — 收到 reviewer subagent 或人类 reviewer 反馈时
Closing skills
verification-before-completion — 声称工作完成/已修好/准备审查前
finishing-a-development-branch — 验证与审查通过后收尾分支
Meta 侧流(改 skill / 改 agent 本身时才用)
writing-skills — 新建或修改 .agents/skills/ 下任何 SKILL.md 时
Planning Gate (HARD GATE)
Before writing ANY code, check these triggers:
Trigger (any ONE suffices):
- Affected file count ≥ 3
- Any change to public API surface — covers all of:
- Module root
.h (e.g., GrowingAutotracker/GrowingAutotracker.h, GrowingTracker/GrowingTracker.h)
GrowingTrackerCore/Public/*.h
GrowingAutotrackerCore/Public/*.h
- Any
Modules/*/Public/*.h
GrowingAnalytics.podspec public_header_files 声明或 Package.swift targets
If triggered:
- Invoke
writing-plans → save plan to docs/plans/YYYY-MM-DD-<feature>.md
- Invoke
plan-document-review → dispatch reviewer subagent
- Show user plan path + reviewer summary, ask for confirmation
- Do NOT touch any source file until user confirms
Required sections in every plan (missing any = incomplete):
- 影响文件列表
- 公开 API 变更(无则填"无")
- 数据协议变更(无则填"无")
- 需同步修改的文档(无则填"无")
Rationalizations (all invalid)
| Excuse | Reality |
|---|
| "改动很简单,不需要规划" | 简单改动也有影响面,规划 2 分钟没有例外 |
| "先改完再补规划" | 规划的价值在事前对齐 |
| "用户没要求规划" | 触发条件满足即强制 |
| "只改内部实现,不影响公开 API" | 内部实现影响 ≥3 文件同样触发 |
| "步骤很清楚,脑内过 checklist 就行" | TaskCreate 落盘 30 秒,无例外 |
Workflow Routing
主线(按序,跳过条件见每条 skill 的 description):
- 模糊需求 →
brainstorming → docs/specs/ → 用户确认
- 读领域知识(AGENTS.md 模块索引 + 源码 + 近期 commit)
- Planning Gate(见上节)→
writing-plans → plan-document-review → 用户确认
- 任务 ≥3 且独立 →
subagent-driven-development;否则直接实施
verification-before-completion
sdk-code-review(有 plan 走模式 A,无 plan + <3 files + 不动公开 API 走模式 B)
- 反馈 →
receiving-code-review → 回到 5
finishing-a-development-branch
旁路(随时插入):
- 改核心路径 →
test-driven-development
- build/test/runtime 失败 →
systematic-debugging
完整流程图见 docs/agents-skills-flow.md。
Red Flags — STOP if you catch yourself thinking these
| Thought | Reality |
|---|
| "这是个简单问题,直接回答就行" | 问题也是任务,先查 skill |
| "小改动不用走流程" | Planning Gate 的触发条件本身就是"小改动"判定器 |
| "用户没要求,跳过吧" | skill 适用性由触发条件判定,不由用户触发 |
| "checklist 我记住了,不用 TaskCreate" | 落盘是约束力,立刻 TaskCreate |
| "需求我看懂了,不用 brainstorming" | 看懂字面 ≠ 规格闭合 |
Remember: The persona defines who you are. Skills define how you work.