edgeone-makers-agents
Pass
Audited by Gen Agent Trust Hub on Aug 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a documentation and framework-specific implementation guide for EdgeOne Makers. It contains skeleton code and best practices for routing, environment variable management, and tool integration.
- [SAFE]: Environment variable handling strictly follows the platform's 'Iron Rule', mandating the use of
context.envorctx.envto prevent process-level pollution and accidental exposure viaprocess.envoros.environ. - [SAFE]: Remote code execution and sandboxing patterns (e.g.,
context.sandbox.runCode) are documented as platform-injected capabilities intended for secure, isolated execution of user-generated code within a managed environment. - [SAFE]: External dependencies referenced (e.g.,
@anthropic-ai/claude-agent-sdk,langchain,crewai,openai) are well-known, industry-standard libraries for AI development. - [SAFE]: The documentation includes critical safety instructions, such as capping agent loops (
maxTurns), handling cancellation signals (AbortSignal), and sanitizing tool outputs to prevent common AI agent failure modes.
Audit Metadata