The Agent Skills Directory

[SAFE]: The skill provides legitimate technical guidance for LayerZero V2, including smart contract development and client-side integration examples.
[EXTERNAL_DOWNLOADS]: The skill references official LayerZero Labs packages (e.g., @layerzerolabs/oapp-evm, @layerzerolabs/oft-evm) and GitHub repositories. These are well-known and established services within the blockchain ecosystem.
[DATA_EXPOSURE]: Deployment scripts and client examples correctly utilize environment variables (e.g., process.env.PRIVATE_KEY) for sensitive credentials, adhering to standard security best practices for avoiding hardcoded secrets.
[INDIRECT_PROMPT_INJECTION]: The skill defines patterns for processing cross-chain messages via the _lzReceive function. These messages represent untrusted external data. While the examples demonstrate basic decoding and state updates, they do not expose any dangerous capabilities (like eval or file system access). Developers utilizing these patterns are responsible for implementing application-specific validation and sanitization of message payloads.
Ingestion points: _lzReceive function defined in SKILL.md, templates/oapp-starter.sol, and examples/send-message/README.md.
Boundary markers: None explicitly used for the payload data.
Capability inventory: No dangerous functions (subprocess, exec, file-write) are present in the provided scripts.
Sanitization: Standard abi.decode is used; further logic-level validation is omitted for brevity in documentation.

layerzero