The Agent Skills Directory

[SAFE]: The skill documentation and instructions are consistent with its stated purpose of providing a technical reference for Eino orchestration. No malicious patterns, such as prompt injection, data exfiltration, or unauthorized code execution, were identified.\n- [SAFE]: All external references and Go import paths point to official repositories within the 'cloudwego' GitHub organization, which is the verified author of the framework. There are no signs of typosquatting or redirection to untrusted domains.\n- [SAFE]: The instructions for the agent emphasize best practices for stability and security, including explicit node compilation and the proper closing of stream readers to prevent resource leaks.\n- [SAFE]: No obfuscation, hidden instructions, or credential exposure patterns (such as hardcoded API keys) were found in any of the skill's reference files.

eino-compose