agent-v3-security-architect
Fail
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- EXTERNAL_DOWNLOADS (HIGH): The
post_executionhook triggersnpx agentic-flow@alpha. This command downloads and executes code from the npm registry at runtime. Theagentic-flowpackage is not from a trusted organization, posing a risk of remote code execution if the package is malicious or compromised. - COMMAND_EXECUTION (MEDIUM): The
post_executionhook interpolates the$TASKenvironment variable directly into a shell command:--task "Security Architecture: $TASK". Without proper sanitization or escaping, an attacker providing a crafted task description containing shell metacharacters (e.g.,;,`,$(...)) could execute arbitrary commands on the host system. - INDIRECT PROMPT INJECTION (LOW): The skill possesses an ingestion surface where external data (
$TASK) is used in a high-capability hook (shell execution). There are no boundary markers or sanitization logic present to prevent instructions embedded within the task data from influencing the shell environment.
Recommendations
- AI detected serious security threats
Audit Metadata