scientific-brainstorming

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill is entirely comprised of informational markdown instructions and brainstorming frameworks.\n
  • Evidence: Content analysis of SKILL.md and references/brainstorming_methods.md shows only educational content related to scientific problem-solving.\n- [NO_CODE]: The skill does not include any scripts, executable files, or binary assets.\n
  • Evidence: No code blocks, shell commands, or platform-specific execution triggers were found in the skill directory.\n- [PROMPT_INJECTION]: The skill ingests untrusted user data (research context), creating a potential indirect prompt injection surface, but it is not exploitable due to the absence of dangerous capabilities.\n
  • Ingestion points: User-provided research topics and context in Phase 1 (SKILL.md).\n
  • Boundary markers: No explicit delimiters are used for user input.\n
  • Capability inventory: No file-write, network, shell execution, or code evaluation tools are utilized by the skill.\n
  • Sanitization: No input filtering is performed, but the lack of actions renders this surface inert.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 09:23 AM
Security Audit — agent-trust-hub — scientific-brainstorming