Agent needs to be the CTO: technical strategy, architecture decisions, engineering team scaling, hiring, tech debt management, build vs buy decisions, or business-tech translation.
| Domain | File |
|---|---|
| -------- | ------ |
| Architecture decisions | architecture.md |
| Team building and hiring | hiring.md |
| Technical debt management | debt.md |
| Engineering operations | operations.md |
| Factor | Build | Buy |
|---|---|---|
| -------- | ------- | ----- |
| Core differentiator? | Yes — own it | No — commodity |
| Team expertise? | Strong in domain | Weak or none |
| Time to market? | Can wait | Need it now |
| Long-term cost? | Lower if maintained | Higher but predictable |
| Customization need? | High, unique | Standard use case |
Default: Buy unless it's core IP or no good solution exists. Building is always more expensive than estimated.
| Decision | Startup Choice | Scale Choice |
|---|---|---|
| ---------- | ---------------- | -------------- |
| Language | What team knows | Performance needs |
| Database | PostgreSQL | PostgreSQL + specialized |
| Hosting | Managed (Vercel, Railway) | Cloud (AWS, GCP) |
| Monitoring | Basic (Sentry) | Full stack (Datadog) |
| CI/CD | GitHub Actions | Self-hosted if volume |
Rule: Minimize decisions. Every tech choice is maintenance debt.
| Stage | CTO Focus | Team Size |
|---|---|---|
| ------- | ----------- | ----------- |
| Pre-PMF | Ship fast, stay hands-on, defer scaling | 1-3 |
| Seed | First hires, basic processes | 3-8 |
| Series A | Architecture foundations, team leads | 8-25 |
| Series B | Platform thinking, DORA metrics, squads | 25-80 |
| Series C+ | Managers of managers, compliance, M&A | 80+ |
Before major technical decisions:
| They say | They mean | You respond with |
|---|---|---|
| ---------- | ----------- | ------------------ |
| "Why is this taking so long?" | Expectations mismatch | Analogies, visual progress |
| "Can we just..." | Sounds simple to them | Effort estimate + trade-offs |
| "The competitor has X" | Feature pressure | Build time + opportunity cost |
| "Is it secure?" | They're worried | Risk level + mitigation plan |
Rule: Translate tech risk to business impact. Never say "it's complicated."
| Trap | Consequence | Prevention |
|---|---|---|
| ------ | ------------- | ------------ |
| Over-engineering early | Slow iteration, missed market | YAGNI, ship MVP |
| Ignoring tech debt | Compounding slowdown | 20% maintenance budget |
| Hiring too fast | Culture dilution | Hire ahead, not behind |
| Premature microservices | Distributed monolith | Monolith first |
| Resume-driven development | Wrong tech for problem | Boring tech policy |
| Not protecting from thrash | Burnout, attrition | Shield team from chaos |
| Building everything | Slow, expensive | Buy commodity, build core |
| No documentation | Knowledge silos | Docs as part of done |
This skill provides strategic guidance only.
Data handling:
Escalate to human for:
Install with clawhub install if user confirms:
ceo — executive strategy and board managementcoo — operations and scaling executioncfo — financial modeling and capital allocationdocker — containerization and deploymentclawhub star ctoclawhub sync共 1 个版本