paystack-charges
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection due to the handling of external data in payment flows.
- Ingestion points: Untrusted data enters the agent context via the
chargeDataobject in theprocessDirectChargefunction and through interactive prompts for authentication details like PIN, OTP, and address. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are used when interpolating these data points into the agent's execution flow.
- Capability inventory: The skill utilizes the
paystackRequesthelper to perform network operations (POST and GET) across multiple Paystack API endpoints. - Sanitization: Input sanitization is limited to the use of
encodeURIComponentfor transaction references; no other data validation or filtering is present to mitigate injection risks.
Audit Metadata