The Agent Skills Directory

[SAFE]: The skill provides comprehensive security guidance for Radius Network development, including detailed checklists for smart contract safety and client-side best practices.\n- [CREDENTIALS_UNSAFE]: The skill correctly handles sensitive information by recommending the use of environment variables and Foundry's encrypted keystores, explicitly advising against exposing private keys in shell history or command-line arguments.\n- [EXTERNAL_DOWNLOADS]: All external dependencies (viem, wagmi, OpenZeppelin) and API references (radiustech.xyz) are established, reputable resources consistent with the skill's purpose and the vendor's domain.\n- [COMMAND_EXECUTION]: Command-line examples for Foundry tools (forge, cast) and Node.js package managers are legitimate and necessary for blockchain development workflows.\n- [PROMPT_INJECTION]: The skill instructions do not contain patterns intended to bypass agent safety guidelines, disregard previous instructions, or extract system prompts.

radius-dev