injection-checking

Pass

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill acts as a documentation repository and routing guide for identifying and testing injection vulnerabilities. The content is purely informational and provides a framework for security research.
  • [SAFE]: Homoglyph characters identified in the text (e.g., the Cyrillic 'а' in "аpple.com") are utilized exclusively as educational examples in a section dedicated to explaining Punycode and homograph attacks. They are not used to obfuscate the skill's operations or to deceive the agent.
  • [SAFE]: References to sensitive system files (e.g., /etc/passwd) and command execution primitives (e.g., id, whoami) are presented as standard payloads for vulnerability assessment and are not executed by the skill itself.
  • [SAFE]: The skill does not perform network operations, access local credentials, or install external dependencies. The link structure points to local relative documentation files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 21, 2026, 01:23 PM
Security Audit — agent-trust-hub — injection-checking