brainstorming

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests untrusted user input and interpolates it into prompts for secondary subagents.
  • Ingestion points: User-controlled content is supplied via the $brainstorming variable as described in SKILL.md.
  • Boundary markers: The instructions do not define clear delimiters (such as XML tags) or safety instructions for subagents to ignore potentially malicious directives embedded within the user input.
  • Capability inventory: The skill utilizes subagent spawning capabilities via fork_turns: "none" in SKILL.md to execute the dynamically generated prompts.
  • Sanitization: There is no evidence of filtering, escaping, or validation of the user input before it is passed to the subagent logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 12:11 PM
Security Audit — agent-trust-hub — brainstorming