lov-bp-polish
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled Python script (
scripts/audit_bp.py) to perform structural and evidence hygiene checks on business plan outlines. The execution is performed with locally resolved paths within the skill's environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided content from documents such as PDF, PPTX, and Markdown files during its 'adversarial review' phase. The workflow lacks explicit boundary markers (e.g., XML tags or specific delimiters) or instructions for the agent to ignore embedded commands within the analyzed documents, which could allow a malicious document to influence agent behavior.
- Ingestion points: The skill reads the business plan outline (
outline.md) and presentation files (PDF/PPTX) into the agent's context for evaluation. - Boundary markers: No specific delimiters or safety instructions are defined in the SKILL.md workflow to isolate user-provided content from system instructions.
- Capability inventory: The skill possesses capabilities for file system read/write operations and the ability to execute the local Python interpreter.
- Sanitization: The input document text is processed without explicit sanitization or filtering for potential injection patterns before being reviewed by the language model.
Audit Metadata