eu-sparql-search
Fail
Audited by Gen Agent Trust Hub on Jul 14, 2026
Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The instructions explicitly direct the agent to subvert platform security controls and tool restrictions.
- Evidence: "This bypasses all web_fetch permission restrictions and works unconditionally"
- Evidence: "Do NOT use web_fetch for Cellar URLs... Use curl in bash_tool instead."
- [COMMAND_EXECUTION]: The skill instructs the agent to use the
bash_toolto execute shell commands and Python scripts for network access, despitebash_toolbeing restricted by the skill's frontmatterallowed-toolsconfiguration. It also provides a code snippet to disable SSL certificate verification (ssl.CERT_NONE), facilitating insecure network connections and potential man-in-the-middle attacks.
Recommendations
- AI detected serious security threats
Audit Metadata