scope-audit
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of processing untrusted data from external scope documents.
- Ingestion points: Processes user-provided text, pasted snippets, and uploaded files (SOWs, RFPs, client emails) through the
ScopeAuditorscript and the agent's qualitative analysis steps. - Boundary markers: Absent. The instructions do not define delimiters or provide specific prompts to ignore malicious instructions embedded within the processed documents.
- Capability inventory: The agent executes a local Python script (
scripts/scope_audit_engine.py) for pattern detection and provides qualitative domain judgment and recommendations based on the document content. - Sanitization: No sanitization, escaping, or validation of the input text is performed before it is integrated into the agent's context for interpretation.
Audit Metadata