alipay-payment-skill
Fail
Audited by Snyk on Nov 15, 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 documented Alipay payment workflow with pinned, integrity-checked tooling and user-confirmed feedback, but its 402 flow automatically sends a fulfillment receipt to the vendor after resource retrieval, which the user is not asked to approve and which the SKILL.md does not mention.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (low risk: 0.10). The required 402 workflow fetches an arbitrary user-supplied resource URL through
alipay-botand reads the returned resource body andPayment-Neededheader, which are outsider-controlled text the agent must actively request, so this is low risk rather than monitored-feed exposure.
MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).
- Direct money access detected (high risk: 0.70). The skill operates Alipay cashier payments and HTTP 402 payments (lines 3, 43-44). It runs payment commands via the alipay-bot CLI or MCP submit, and a 402 payment runs immediately without a further confirmation (line 44). That is a direct financial operation that charges or pays a real order. For control, the skill documents a wallet lifecycle handoff to
alipay-authenticate-wallet(line 47) and says payment commands handle their own wallet readiness (line 50). It does not describe how the service or execution layer authenticates or authorizes the payment route. The evidence offered is instruction-level only: STOP boundaries, no automatic retries, and not retrying after errors. Those are agent-followed rules, not service-enforced access checks. A named wallet skill does not show service authentication, and we should not infer it. No qualifying independent control is established for the direct payment route, so the tier is high.
Issues (3)
E004
CRITICALPrompt injection detected in skill instructions.
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
W009
MEDIUMDirect money access capability detected (payment gateways, crypto, banking).
Audit Metadata