business-plan-advisor
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is designed to provide structured consultation for business planning, using local reference materials and utility scripts to assist users.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided business plan documents for review and validation. While this represents a potential attack surface for indirect prompt injection, the risk is minimized by the use of structured validation scripts (
scripts/validate_business_plan.py) that rely on regular expressions for data extraction rather than exposing the agent directly to unformatted external content. Evidence:\n - Ingestion points: User-provided business plan documents referenced in
SKILL.mdworkflows.\n - Boundary markers: None explicitly defined in the prompt instructions.\n
- Capability inventory: Local file reading and command execution via provided scripts (
validate_business_plan.py,financial_calculator.py). No network or sensitive credential access.\n - Sanitization: The validation script performs pattern-based extraction for numbers and sections, which acts as a primitive filter against injection content.
Audit Metadata