The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several TypeScript scripts (e.g., execute-minting.ts, reserve-collateral.ts, xrp-payment.ts) designed to execute blockchain transactions on the Flare and XRP Ledger networks. These scripts require sensitive inputs like private keys or seeds, which are managed via environment variables or designated placeholders (PUT_SEED_HERE). Most write-capable scripts implement a DRY_RUN safety mechanism that prevents execution unless explicitly enabled by the user.
[EXTERNAL_DOWNLOADS]: The documentation and scripts reference external dependencies and APIs, including the Flare Data Connector (FDC) verifier and official Flare contract repositories. These resources are part of the flare-foundation ecosystem and standard blockchain development tools (e.g., ethers, xrpl, viem).
[PROMPT_INJECTION]: The skill describes an architecture that interacts with external data such as XRPL payment memos and FDC attestation proofs, which constitutes an indirect prompt injection surface. Ingestion points: SKILL.md identifies FDC attestation payloads, XRPL payment references, and verifier responses as external data sources. Boundary markers: The skill explicitly instructs developers to treat this data as untrusted and never pass it into LLM prompts. Capability inventory: Scripts like execute-minting.ts and reserve-collateral.ts perform network requests and blockchain transactions across all relevant files. Sanitization: The instructions emphasize strict binary decoding and validation of all external data before use.

flare-fassets