重要前提
安装AI Skills的关键前提是:必须科学上网,且开启TUN模式,这一点至关重要,直接决定安装能否顺利完成,在此郑重提醒三遍:科学上网,科学上网,科学上网。查看完整安装教程 →
gtm-developer-ecosystem by github/awesome-copilot
npx skills add https://github.com/github/awesome-copilot --skill gtm-developer-ecosystem通过生态计划、社区和合作伙伴关系,构建并扩大由开发者主导的采用。专注于真正推动采用的因素,而非虚荣指标。
触发场景:
适用背景:
模式: 在开发者平台上运营生态。领导层争论:向所有人开放市场,还是精选以保证质量?
质量控制派: "我们需要设立准入门槛。否则会充斥 SEO 垃圾、低质量集成,损害品牌声誉。" 开放派: "开发者会绕过守门人。网络效应比质量控制更重要。"
决策: 选择开放。质量担忧确实存在,但我们下注:控制力应来自发现和信任层,而非提交时的准入门槛。
我们构建的替代准入门槛的方案:
结果: 网络效应胜出。发布了数千个集成。质量通过实际使用情况浮现,而非由我们预先决定。
决策框架:
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
常见错误: 默认选择精选,因为"我们需要质量控制"。这在你有 10 个合作伙伴时有效。当达到 100 个以上时,你自己就成了瓶颈。应转而构建发现和信任系统。
模式: 大多数开发者项目追求短期效益。更好的方法是:建立长期人才输送管道。
第一年:大学合作
第二年:学生社区与认证
第三年:职业桥梁
为何有效: 学生在 5-10 年后会成为企业买家。你在他们拥有购买力之前就建立了品牌忠诚度。
常见错误: 将学生视为即时收入来源。他们不是。他们是未来的企业决策者。
阶段 1:认知
阶段 2:入门
阶段 3:集成
阶段 4:生产环境
阶段 5:拥护
关键指标:
常见错误: 测量虚荣指标(注册数、下载量)而非真实参与度(API 调用量、生产环境部署数)。
层级 1:快速入门(快速获得价值)
层级 2:指南(解决实际问题)
层级 3:参考(完整的 API 文档)
层级 4:概念性(理解系统)
大多数开发者需要: 首先是层级 1,然后是层级 2。很少有人阅读层级 4。
常见错误: 从层级 3(全面的 API 参考)开始。开发者首先想要快速获得成果。
社区(异步,可扩展):
支持(同步,成本高):
如何分流:
社区优先:
升级到支持的情况:
常见错误: 为所有人提供白手套支持。这无法扩展。应构建能够自我帮助的社区。
层级 1:集成合作伙伴(自助服务)
层级 2:战略合作伙伴(共同开发)
不要过度分级。 2 个层级足够。更多会造成混淆。
如果低质量合作伙伴加入,品牌损害风险高吗?
├─ 是(受监管、安全) → 精选
└─ 否 → 继续...
│
你能规模化人工审核吗?
├─ 否(成百上千) → 开放 + 发现系统
└─ 是(数十个) → 精选
这是常见问题吗?
├─ 是 → 社区(论坛、Slack、文档)
└─ 否 → 继续...
│
请求者是付费客户吗?
├─ 是 → 支持(邮件、专用)
└─ 否 → 社区(带升级路径)
1. 在产品市场匹配前构建生态
2. 没有开发者成功团队
3. 文档质量差
4. 平等对待所有开发者
5. 没有集成质量标准
6. 只测量虚荣指标
7. 开发者布道师缺乏技术深度
开放生态清单:
开发者旅程指标:
文档层级:
合作伙伴分级:
学生项目时间线:
基于在多家平台公司构建开发者生态的经验,包括开放与精选市场的决策、学生项目开发(构建人才输送管道的三年规划)以及合作伙伴生态增长。并非理论——来自实际推动平台采用和多年品牌忠诚度的开发者生态构建模式。
每周安装量
175
代码仓库
GitHub 星标数
26.9K
首次出现
6 天前
安全审计
安装于
codex162
opencode162
gemini-cli161
cursor160
github-copilot158
kimi-cli158
Build and scale developer-led adoption through ecosystem programs, community, and partnerships. Focus on what actually drives adoption, not vanity metrics.
Triggers:
Context:
The Pattern:
Running ecosystem at a developer platform. Leadership debate: Open the marketplace to anyone, or curate for quality?
Quality control camp: "We need gatekeeping. Otherwise we'll get SEO spam, low-quality integrations, brand damage."
Open camp: "Developers route around gatekeepers. Network effects matter more than quality control."
The decision: Went open. Quality concerns were real, but we made a bet: control comes from discovery and trust layers, not submission gatekeeping.
What We Built Instead of Gatekeeping:
Result: Network effects won. Thousands of integrations published. Quality surfaced through usage, not through us deciding upfront.
Decision Framework:
Common Mistake:
Defaulting to curated because "we need quality control." This works when you have 10 partners. At 100+, you become the bottleneck. Build discovery and trust systems instead.
The Pattern:
Most developer programs optimize for quick wins. Better approach: Build long-term talent pipeline.
Year 1: University Partnerships
Year 2: Student Community & Certification
Year 3: Career Bridge
Why This Works:
Students become enterprise buyers 5-10 years later. You're building brand loyalty before they have purchasing power.
Common Mistake:
Treating students as immediate revenue. They're not. They're future enterprise decision-makers.
Stage 1: Awareness
Stage 2: Onboarding
Stage 3: Integration
Stage 4: Production
Stage 5: Advocacy
Metrics That Matter:
Common Mistake:
Measuring vanity metrics (sign-ups, downloads) instead of real engagement (API calls, production deployments).
Tier 1: Quick Starts (Get to Value Fast)
Tier 2: Guides (Solve Real Problems)
Tier 3: Reference (Complete API Docs)
Tier 4: Conceptual (Understand the System)
Most developers need: Tier 1 first, then Tier 2. Very few read Tier 4.
Common Mistake:
Starting with Tier 3 (comprehensive API reference). Developers want quick wins first.
Community (Async, Scalable):
Support (Sync, Expensive):
How to Route:
Community first:
Escalate to support when:
Common Mistake:
Providing white-glove support to everyone. Doesn't scale. Build community that helps itself.
Tier 1: Integration Partners (Self-Serve)
Tier 2: Strategic Partners (Co-Development)
Don't over-tier. 2 tiers is enough. More creates confusion.
Is brand damage risk high if low-quality partners join?
├─ Yes (regulated, security) → Curated
└─ No → Continue...
│
Can you scale human review?
├─ No (hundreds/thousands) → Open + discovery systems
└─ Yes (dozens) → Curated
Is this a common question?
├─ Yes → Community (forum, Slack, docs)
└─ No → Continue...
│
Is requester paying customer?
├─ Yes → Support (email, dedicated)
└─ No → Community (with escalation path)
1. Building ecosystem before product-market fit
2. No developer success team
3. Poor documentation
4. Treating all developers equally
5. No integration quality standards
6. Measuring only vanity metrics
7. Developer advocates with no technical depth
Open ecosystem checklist:
Developer journey metrics:
Documentation hierarchy:
Partner tiers:
Student program timeline:
Based on building developer ecosystems at multiple platform companies, including the open vs curated marketplace decision, student program development (3-year arc building talent pipeline), and partner ecosystem growth. Not theory — patterns from building developer ecosystems that actually drove platform adoption and multi-year brand loyalty.
Weekly Installs
175
Repository
GitHub Stars
26.9K
First Seen
6 days ago
Security Audits
Gen Agent Trust HubPassSocketPassSnykPass
Installed on
codex162
opencode162
gemini-cli161
cursor160
github-copilot158
kimi-cli158
agent-browser 浏览器自动化工具 - Vercel Labs 命令行网页操作与测试
169,700 周安装
Next.js全栈测试方案:Neon数据库分支、Playwright、集成与单元测试完整指南
80 周安装
SnapDOM:快速无依赖的DOM转图像库,支持SVG/PNG/JPG/WebP格式导出
80 周安装
MVP案例构建器 - AI驱动的体育奖项评选与数据分析工具 | 专业论证生成
82 周安装
Vuex Vue2 状态管理教程:安装、核心概念与实战指南
82 周安装
Orval:OpenAPI/Swagger 转 TypeScript 客户端代码生成器 | 支持 React Query、Vue、Angular
81 周安装
移动应用开发技能指南 - 掌握iOS/Android开发、Flutter框架与移动端最佳实践
58 周安装