divorce-practice-stephane-boghossian
Pass
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data, creating an attack surface for indirect prompt injection. \n- Ingestion points: The workflow in
SKILL.md(specifically Mode 0 and Mode 4) requires the agent to process "uploaded intake forms," "tax returns, deeds, account statements," and "communication / message-thread analysis (e.g. an exported chat history)." \n- Boundary markers: The instructions do not specify the use of delimiters or provide instructions for the agent to ignore potentially malicious commands embedded within the documents being analyzed. \n- Capability inventory: The skill identifies and interacts with various case facts but consists solely of instructional markdown; it lacks executable code, shell scripts, or defined network capabilities. \n- Sanitization: No sanitization or validation steps are defined for the content extracted from external files before it is processed by the agent. \n- [NO_CODE]: The skill consists entirely of instructional text and metadata, containing no executable scripts or binaries.
Audit Metadata