accept-agent-payments
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources including developer documentation, marketplace listings, and raw HTTP 402 Payment Required responses.
- Ingestion points: Fetches content from external URLs (developers.circle.com, agents.circle.com) and processes dynamic metadata from inspected services.
- Boundary markers: Includes explicit instructions to treat fetched content as untrusted and to prioritize system instructions over data-embedded instructions.
- Capability inventory: Generates server-side Node.js code and constructs shell commands for testing (curl, circle CLI).
- Sanitization: Implements a dedicated Command Safety section that validates input formats (URLs, addresses, amounts) and rejects shell metacharacters like semicolons or backticks.
Audit Metadata