Optimizes and validates agent prompt files by enforcing prompt engineering best practices and security patterns.
Agent Prompt Enhancer is a comprehensive utility designed to audit and improve AI agent definitions for platforms like Claude Code and OpenCode. It analyzes agent markdown files against 30 distinct patterns, including frontmatter accuracy, structural hierarchy, and instruction effectiveness. By enforcing the 'Principle of Least Privilege' for tool use and the 'Intern Test' for descriptions, it ensures agents are reliable, secure, and easy to invoke. The skill provides actionable reports and can automatically fix high-certainty issues like unrestricted tool access or weak constraint language, streamlining the development of high-performance AI agents.
Key Features
01Security auditing for tool permissions and input validation
021 GitHub stars
0330+ automated pattern checks for prompt engineering excellence
04Automated fixes for frontmatter and instruction strength
05Validation of structural hierarchy (Role, Workflow, Constraints)
06Cross-platform support for Claude Code, OpenCode, and Codex formats
Use Cases
01Auditing existing agent libraries for security vulnerabilities and token bloat
02Improving agent reliability by converting weak suggestions into mandatory constraints
03Standardizing agent descriptions to ensure correct invocation during complex workflows
