GitHub · TypeScript
shellhaki/hstack
Collection of agent skills where a small, cheap model does the engineering (spec, plan, tests, code) and a frontier model directs it like a senior engineer: intake, checkpoint approvals, final review. 35% cheaper than the frontier alone at equal quality in the first benchmark.
- Stars
- 7
- Forks
- 0
- Stars gained in 24 hours
- Not enough history yet
- Forks gained in 24 hours
- Not enough history yet
Topics: agent-skills, ai-agents, cost-optimization, llm, muse-code
Latest data point: Nov 21, 2026 · 7 stars · 0 forks