setup-bayarlah
Pass
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for the agent to help build systems that ingest external data via payment success callbacks (notify_url) and Event Order Notifications. This creates a potential surface for indirect prompt injection if the incoming data contains malicious instructions that are not properly sanitized before being presented back to an agent or user.
- Ingestion points: Webhook listeners for
notify_urland transaction push notifications. - Boundary markers: The skill explicitly instructs developers to verify signatures over the raw body before parsing JSON, which provides a strong technical boundary.
- Capability inventory: The skill does not contain self-executing scripts but provides logic for file system persistence (terminal keys) and network requests to the payment provider.
- Sanitization: The instructions mandate strict signature verification and use of constant-time comparisons to prevent timing attacks and unauthorized data processing.
- [SAFE]: The skill demonstrates high security awareness by explicitly advising against the use of an unofficial npm package ('bayarlah' by 'razmans') to avoid supply chain risks. It also correctly advises storing all cryptographic keys and secrets server-side rather than in frontend code or static environment variables.
Audit Metadata