The Agent Skills Directory

[SAFE]: The skill is composed of instructional markdown and reference documentation for auditing authorization logic. No executable scripts, binaries, or automated installation commands are present.
[SAFE]: The frontmatter configuration specifies allowed-tools (Read, Grep, Glob, Bash) which are restricted to standard file system search and analysis utilities consistent with a code-auditing persona.
[SAFE]: The code snippets included in the reference documentation are categorized into 'bad' and 'safe' examples to guide the AI agent's detection logic. These are for pattern matching and do not constitute functional code for the skill itself.
[SAFE]: No obfuscation, hardcoded credentials, network exfiltration patterns, or persistence mechanisms were detected. All external references are to legitimate CVEs or established security documentation.

wrdn-authz