agent-modernizer

To audit Claude Code plugin agent definitions, validate frontmatter fields including name, description, model, and color, along with trigger conditions. This skill applies structured, checklist-based validation to ensure existing agent files meet current standards.

Rewriting verbose agent system prompts is done by converting them into concise, decision-driven definitions with clear boundaries. This skill applies a structured, frontmatter-first rewrite process to enforce plugin guidelines and best practices.

Batch validation of multiple agent files in a plugin is supported. You can apply the same frontmatter and system-prompt standards across multiple agent definitions simultaneously to align them with current plugin guidelines.

Required frontmatter fields for a valid Claude Code agent definition include name, description, model, and color. The audit process verifies these fields along with trigger conditions to ensure full standards compliance.

Creating a new Claude Code agent following best practices involves using a structured, frontmatter-first rewrite process. This skill applies checklist-based validation and guardrails to output formatting, ensuring new agents meet current standards.

Agent definition validation failures occur when frontmatter fields like name, description, model, or color are missing or misconfigured. The audit process checks these fields and trigger conditions against current standards using structured checklists.