The Agent Skills Directory

[SAFE]: The skill consists of a collection of markdown rule files and reference documents designed to guide an AI agent through a structured audit of agentic UI/UX features. It uses standard system tools like ripgrep (rg) and git to analyze the local workspace for specific code patterns (e.g., checking for the presence of approval gates, progress indicators, or tool parity).
[SAFE]: No evidence of malicious behavior was found. The skill does not attempt to access sensitive credentials (~/.ssh, ~/.aws, etc.), perform unauthorized network operations, or exfiltrate data. All shell commands are scoped to searching the user's provided source code for auditing purposes.
[SAFE]: The skill uses clear, human-readable instructions and does not employ obfuscation techniques like Base64 encoding, hidden Unicode characters, or homoglyph substitutions.
[SAFE]: There are no remote code execution patterns or unverifiable external dependencies. The skill operates entirely on the provided templates and existing system utilities.

ax-audit