paymob-integration

Fail

Audited by Snyk on Nov 14, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.30). The skill is a legitimate Paymob integration guide, but its instructions to strip AI authorship markers from the merchant's commits and to avoid flagging AI-written work hide the agent's own activity from the repository owner.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (low risk: 0.10). The required workflow reads only first-party Paymob docs, the merchant's own code, and Paymob's own callbacks, with community-forum and transaction-data reads being optional and on-demand, so the closest exposure is an agent-initiated fetch rather than a monitored outsider feed.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

  • Direct money access detected (medium risk: 0.30). Direct financial operations are in scope. The skill offers live-account actions that move or authorize money: creating intentions and payment links (line 14, line 119), refunds, voids, captures, and settlements via the MCP server or API (line 299, line 14), and charges through checkout intentions (line 236). These are direct operations, not indirect ones. The skill's own content is guidance, but Phase 4 and the MCP section offer to run these actions on the merchant's real account, so they are operational routes.

Control: the MCP and API routes authenticate with the merchant's own Paymob API credentials (line 14). The evidence standard treats validated service API credentials as qualifying access control, and the skill documents that Paymob's service rejects invalid credentials: "401 Unauthorized on intention create" from a wrong or expired Secret Key (line 316). That shows service-side validation, not just agent behavior. The live-action safety rules (lines 21–23) are agent-followed confirmations and do not count as independent control, but they are not needed to establish the credential gate. Because a qualifying control is established for the direct routes, the severity is medium, not high. The gap is that the skill does not document an explicit independent check for each MCP write tool beyond the shared credential, so the rating relies on that shared authentication.

Issues (3)

E004
CRITICAL

Prompt injection detected in skill instructions.

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

W009
MEDIUM

Direct money access capability detected (payment gateways, crypto, banking).

Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 14, 2026, 07:14 PM
Issues
3