agent-architect

To create an AI agent for code reviews, this Skill guides you through an interactive Q&A process to define the agent's scope, necessary permissions, and preferred tools, then generates the required configuration frontmatter for optimal performance.

The best way to configure OpenCode agent triggers and permissions is through a guided conversational process that helps define operational parameters, scope, and persona, ensuring the agent is precisely tuned for its intended use cases.

To fix an AI agent ignoring context or running slowly, you can use this Skill to diagnose performance issues by analyzing the existing configuration and suggesting operational improvements to optimize execution and context handling.

Yes, you can refine an existing OpenCode agent's operational parameters through iterative feedback, adjusting specific triggers, permissions, and skill access without rebuilding the entire configuration from scratch.

Before building an AI agent configuration, you need to define the agent's purpose, scope, triggers, persona, and necessary permissions so the guided process can map these to appropriate skill access and operational modes.

Limitations to tuning AI agent performance through configuration adjustments exist if the initial scope is poorly defined, as the Skill relies on iterative feedback to ensure agents have appropriate skill access and correctly configured operational modes.