brainstorm

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted user input via the $ARGUMENTS variable in SKILL.md to define the feature description.
  • Ingestion points: The <feature description> block in SKILL.md captures user-supplied input.
  • Boundary markers: The input is wrapped in XML-style <feature description> tags, providing basic structural delimitation.
  • Capability inventory: The skill has the capability to write to the local filesystem (docs/brainstorm/), create git branches (@create-branch), and perform codebase research via @codebase-review-agent.
  • Sanitization: No explicit sanitization or 'ignore instructions' guards are applied to the interpolated content before it is processed by subsequent tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 09:05 PM
Security Audit — agent-trust-hub — brainstorm