The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill identifies surfaces where untrusted data from the XRP Ledger (such as transaction Memos or metadata) enters the agent's context.
Ingestion points: references/payments.md includes monitor_account (WebSocket subscription) and decode_memo (transaction lookup) which ingest on-chain data.
Boundary markers: The provided code snippets do not implement specific boundary markers or 'ignore' instructions for data extracted from Memos.
Capability inventory: The skill possesses the capability to construct payment, escrow, and trust line transactions through the xrpl and xrpl-py libraries.
Sanitization: Memos are decoded from hex and parsed as JSON, but the skill does not include specific logic to filter for malicious natural language instructions within those fields.
[EXTERNAL_DOWNLOADS]: The skill references several external resources and network endpoints essential for XRP Ledger operations.
Evidence: The skill connects to well-known infrastructure including the XRPL Testnet (s.altnet.rippletest.net), Mainnet RPC nodes (xrplcluster.com), and official documentation from Ripple (docs.ripple.com). These references are consistent with the skill's primary purpose and target well-known services.

xrpl-payments