The Agent Skills Directory

[SAFE]: The skill is authored by 'aptos-labs' and focuses on providing secure coding patterns for the Aptos blockchain. All instructions and examples align with official security best practices.
[SAFE]: No malicious patterns such as prompt injection, data exfiltration, or unauthorized command execution were detected.
[CREDENTIALS_UNSAFE]: The skill explicitly instructs the agent to avoid hardcoding secrets and warns against reading sensitive files like .env or ~/.aptos/config.yaml, which is a positive security practice.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests user input to generate contracts, it provides extensive rules for the agent to follow (e.g., 'ALWAYS verify signer authority', 'ALWAYS validate inputs') to ensure the generated code is secure. It includes a comprehensive security checklist and references official security guidelines from 'aptos.dev'.

write-contracts