owasp-top-ten-check
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a set of text-based instructions for a security audit. It does not contain any executable scripts, binary files, or automated commands.
- [COMMAND_EXECUTION]: Mentions of potentially dangerous strings (e.g.,
' OR '1'='1) and commands (e.g.,cat /etc/passwd) are used strictly as illustrative examples of what an auditor should look for in a target application. They are not directed at the agent's host environment. - [EXTERNAL_DOWNLOADS]: The skill includes links to well-known and highly trusted security resources, including the OWASP Foundation, MITRE (CWE), and PortSwigger. These are industry-standard references for security professionals.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for untrusted data via
$ARGUMENTS(representing the code or architecture to be audited). However, the skill does not define or invoke any high-privilege tools (such as file writers or network clients) that could be exploited if malicious instructions were present in that data. The assessment is purely diagnostic in nature.
Audit Metadata