evolution-review
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill consists entirely of natural language instructions and evaluation examples. No executable scripts (Python, JavaScript, shell) are included in the skill package.- [SAFE]: The instructions describe a business logic process for evaluating performance data. No patterns of prompt injection, data exfiltration, or malicious behavior were detected during the analysis.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external portfolio data (items, revenue, reach metrics) provided by the user.
- Ingestion points: User-provided portfolio data via chat interaction.
- Boundary markers: None explicitly defined in the instructions to separate data from instructions.
- Capability inventory: No capabilities (file system, network, or subprocess execution) are requested or used by the skill.
- Sanitization: None present.
- Note: The total lack of system capabilities effectively mitigates the risk of indirect prompt injection, as there are no actions for an attacker to hijack.
Audit Metadata