kdcokenny/opencode-workspace
Bundled multi-agent orchestration harness for OpenCode. One install, complete control.
What it solves
It provides a pre-configured, multi-agent orchestration harness for AI-driven software development. Instead of manually configuring individual agents, plugins, and tools, this bundle allows users to deploy a complete environment where specialized agents (like coders, researchers, and reviewers) work together under the orchestrators.
How it works
The system uses a hierarchical architecture where orchestrators (plan and build) delegate tasks to specialists. These specialists are constrained by specific permission boundaries to ensure security and stability. For example, the coder agent has full bash access, while the researcher is restricted to read-only MCP tools for external research.
Who it’s for
Developers who want a ready-made, multi-agent AI workspace for coding, research, and documentation without having to build the orchestration logic from scratch.
Highlights
- Multi-Agent Ecosystem: Includes 4 specialized agents (researcher, coder, scribe, and reviewer) and 2 orchestrators.
- Curation of Tools: Bundles 3 MCP servers (Context7, Exa, GitHub Grep) and various plugins for plan management and Git worktree isolation.
- Integrated Skills: Includes built-in guidelines for implementation planning, code review methodology, and frontend/internal logic philosophies.
- Security Boundaries: Implements strict permission settings, such as denying direct web fetching and restricting bash access to specific agents.
Related
- Project
- Project
- Project
- Project
- Project