OpenAgents Control (OAC)
A toolkit that extends the OpenCode framework so AI coding agents learn a team's existing coding patterns, architecture and standards before writing code, then request approval before making changes.
OpenAgents Control (OAC) is built on OpenCode, an open-source AI coding framework, and adds specialized agents, context management and team workflows on top of it. Agents load a project's own patterns before generating code, so output matches the codebase's conventions instead of generic suggestions that need rewriting.
Agents always propose a plan and wait for approval before executing it, and their behavior is defined in editable markdown files rather than compiled plugins, so teams can change workflows and constraints directly. OAC targets multiple languages, including TypeScript, Python, Go and Rust, and works with several model providers.