popup-cro
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious code, obfuscation, or unauthorized access patterns were found. The skill consists of instructional markdown and local data processing scripts using the Python standard library.
- [INDIRECT_PROMPT_INJECTION]: The scripts ingest user-provided JSON data (popups.json, test.json, roi_data.json), creating a potential attack surface. 1. Ingestion points: scripts/popup_ab_test_calculator.py, scripts/popup_roi_estimator.py, and scripts/popup_strategy_auditor.py via json.load(). 2. Boundary markers: Absent. 3. Capability inventory: None detected across scripts; there are no network operations, shell command executions, or file-writing capabilities. 4. Sanitization: The scripts perform basic data validation and mathematical operations on numeric fields. The absence of dangerous capabilities makes this surface safe.
Audit Metadata