decide

Pass

Audited by Gen Agent Trust Hub on Feb 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the system date command (e.g., date +%Y%m%d-%H%M%S) to generate unique timestamps for session logging and file naming.
  • [EXTERNAL_DOWNLOADS]: The skill uses a web-researcher subagent to retrieve data from external web sources, which is necessary for comparing products, services, or other external decision factors.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes content from external websites. * Ingestion points: External data enters the context via the web-researcher subagent and direct web searches. * Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present when handling retrieved content. * Capability inventory: The skill can execute local system commands (date), read/write files to the /tmp directory, and perform further network operations through subagents. * Sanitization: No sanitization or validation of retrieved external data is performed before it is analyzed and synthesized into recommendations.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 23, 2026, 03:42 PM