Master BP Reviewer
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted user-provided business documents (PDFs, pitch decks) without using explicit boundary markers or delimiters to separate user data from instructions. Ingestion points: Pitch decks and business documents provided by users via the Read tool (SKILL.md). Boundary markers: Absent. Capability inventory: File system read access via the Read tool and file system write access for saving generated reports. Sanitization: Absent.
- [COMMAND_EXECUTION]: The skill utilizes tools to read content from user documents and write review reports to the current working directory. While necessary for its core function, these file system operations represent a capability surface that could be targeted if malicious document content or filenames were to influence the agent's file processing.
Audit Metadata