crypto-payments

Fail

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: CRITICALEXTERNAL_DOWNLOADSDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill is centered around integrating with api.heleket.com and heleket.com. These domains are identified by automated security scanners as phishing sites. Connecting an application to these endpoints to process financial transactions poses a high risk of funds being routed to malicious actors.\n- [DATA_EXFILTRATION]: The integration instructions involve sending transaction details, merchant identifiers, and potentially sensitive merchant metadata to api.heleket.com. Given the phishing classification of the destination, this represents a high risk of sensitive data exfiltration.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a public webhook endpoint (/api/billing/crypto/webhook) that ingests untrusted data from an external provider.\n
  • Ingestion points: The POST endpoint receives JSON payloads representing payment status changes from the external gateway.\n
  • Boundary markers: No specific boundary markers or instructions for the agent to ignore instructions embedded in the payload are present in the processing logic described.\n
  • Capability inventory: The processing logic includes database transactions (db.$transaction), user credit updates (creditAccountPurchasedTokens), and notification triggers (notify).\n
  • Sanitization: The skill advocates for MD5 signature verification and IP allowlisting (specifically targeting 31.133.220.8). While these are intended security gates, MD5 is a cryptographically weak hashing algorithm, and the lack of robust input sanitization creates a vulnerability surface.
Recommendations
  • AI detected serious security threats
  • Contains 4 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 12, 2026, 12:34 AM