Feedback wanted: try the OpenCode Agent Orchestration Kit

[jcarlosrodicio]

Hi everyone!

I’ve published my OpenCode orchestration setup as a public, reproducible repository:

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

I built it because my OpenCode workflow was becoming difficult to maintain: planning, research, design, implementation and review were starting to overlap, and changing models per agent required manually editing configuration files.

The kit provides:

• Specialized agents for routing, research, design, specification, implementation, review and validation.
• Slash commands for planning, features, research, design, implementation, testing and review.
• Bounded workflows with explicit handoffs and human approval points.
• Open Design integration for local, editable product and UI workflows.
• oc-switch, an interactive TUI for changing the default, small/title and per-agent models using OpenCode’s configured catalog.
• Backup-aware installation, upgrades, rollback and validation checks.

I’m looking for hands-on feedback rather than just stars.

If you try it, I’d especially like to know:

1. Was the installation process clear?
2. Did the agent roles and handoffs make sense?
3. Was oc-switch useful for changing model assignments?
4. Was the Open Design integration valuable?
5. What felt confusing, unnecessary or over-engineered?
6. What documentation or workflow is missing?

You can test only one part of the kit; you don’t need to adopt the entire setup.

Please include your OS, OpenCode version, Node version and what you tested. Bugs can go into Issues, while broader workflow feedback can stay in this discussion.

This is an independent configuration and workflow kit for OpenCode, not an official OpenCode project.