@jcarlosrodicio2
Founder, Rcuadrado
1h ago · Agents · Seville

Building a more reliable way to orchestrate AI coding agents with OpenCode

Hi everyone! I'm Juan Carlos, a software engineer from Spain with over 11 years of experience.

Over the past few months, I've been exploring how to give AI coding agents more autonomy without losing control over software quality, security, and engineering decisions.

This led me to build OAK (OpenCode Agent Orchestration Kit), an open-source project that has recently reached 154 GitHub stars.

The idea is to turn OpenCode into a structured software engineering team with:

  • Specialized agents for research, planning, implementation, and review.
  • Bounded autonomous workflows with persistent, recoverable state.
  • Independent code review and deterministic validation.
  • Explicit human approval gates and traceable decisions.
  • Over 1,000 automated tests.

I believe the next challenge in AI-assisted development isn't just making agents more autonomous, but making that autonomy reliable, measurable, and safe.

I'd love to hear how others in this community are approaching this problem.

GitHub: github.com/jcarlosrodicio/opencode-agent-orchestration-kit

AI AgentsAgent OrchestrationOpenCodeOpen SourceHarness EngineeringTools

Comments

No comments yet
Start the conversation.
Log in to join the thread.