customer-journey-stages
Pass
Audited by Gen Agent Trust Hub on Feb 13, 2026
Risk Level: LOWPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION] (LOW): An indirect prompt injection surface exists because the skill ingests untrusted user data to generate structured documentation.
- Ingestion points: The skill requires 'MVP 주제', '고객유형', and '시장 조사 결과' as inputs from the user or external market research data.
- Boundary markers: Absent; the skill does not use delimiters or instructions to prevent the agent from obeying instructions embedded within the input data.
- Capability inventory: The skill includes the capability to write output to the file system at
define/고객경험단계.md. - Sanitization: Absent; there is no validation or escaping of the input data before it is interpolated into the output template.
- [SAFE] (SAFE): No malicious patterns such as remote code execution, hardcoded credentials, data exfiltration, or obfuscation were detected in the skill file.
Audit Metadata