board-of-directors

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use shell commands (mkdir -p) to create a directory structure for managing session metadata and discussion logs within a specific project path.
  • [PROMPT_INJECTION]: The skill is designed to evaluate arbitrary 'proposals' and 'context' data. This creates an indirect prompt injection surface where malicious instructions in a proposal could attempt to influence the board's deliberation, although the 'Chief Security Officer' persona is explicitly tasked with detecting such patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 08:19 PM