gstack
Use Garry Tan's exact Claude Code setup: 15 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA
Star Growth
Overview
gstack is an open-source collection of specialized AI agent tools created by Y Combinator's Garry Tan that transforms Claude Code into a virtual engineering team. Instead of working with generic AI prompts, gstack provides 15+ opinionated specialist roles including CEO (strategic product thinking), Designer (visual review), Eng Manager (architecture decisions), QA Lead (browser testing), Security Officer (OWASP audits), and Release Manager (deployment automation). Each role is accessible through slash commands like /plan-ceo-review, /review, /qa, and /ship. The toolset aims to help technical founders and engineering teams maintain the rigor of a full development team while moving at startup speed. Tan reports shipping over 600,000 lines of production code using these tools, demonstrating their potential for high-velocity development. The project is MIT licensed and designed for founders, first-time Claude Code users, and tech leads who want structured, role-based AI assistance rather than blank-slate interactions.
Deep Analysis
vs bare Claude Code: 23 specialist roles with structured sprint process; vs Cursor: not an IDE but a process layer that turns any coding agent into a virtual team with CEO/design/eng review
⚡ Capabilities
- • Claude Code skill-based virtual engineering team
- • Sprint workflow automation (Think-Plan-Build-Review-Test-Ship)
- • CEO/Eng/Design review skills
- • Browser-based QA testing
- • Security auditing (OWASP/STRIDE)
- • PR creation and release management
- • Office hours for product ideation
- • Retroactive sprint analysis
🔗 Integrations
✓ Best For
- ✓ Solo founders shipping production code with AI
- ✓ Technical CEOs who still want to code
- ✓ Teams wanting structured AI-assisted code review and QA
✗ Not Ideal For
- ✗ Non-Claude Code users
- ✗ Teams with established CI/CD who don't want workflow changes
Languages
Deployment
Pricing Detail
⚠ Known Limitations
- ⚠ Requires Claude Code or compatible agent as host
- ⚠ Quality depends on Claude model capabilities
- ⚠ Markdown-based skills can be fragile
- ⚠ Opinionated workflow may not fit all teams
Pros
- + Provides structured specialist roles instead of generic AI prompts, making interactions more focused and productive
- + Comprehensive workflow coverage from strategic planning to code review, QA testing, and deployment automation
- + Battle-tested by a high-profile user with impressive productivity claims and strong community adoption (52K+ GitHub stars)
Cons
- - Highly opinionated approach may not suit all development workflows or team preferences
- - Requires Claude Code setup and familiarity, limiting accessibility for users of other AI tools
- - May be overly complex for simple projects or developers who prefer minimal tooling
Use Cases
- • Technical founders who want to maintain engineering rigor while shipping code quickly as a solo developer
- • Engineering teams looking to standardize code review, QA, and release processes with AI assistance
- • Claude Code users who want specialized agent roles for different aspects of software development instead of general-purpose prompting