claude
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The provided files (SKILL.md and SKILL.zh.md) contain only documentation and CLI usage examples. No executable scripts, binary files, or hidden payloads are present in the analyzed content.- [COMMAND_EXECUTION]: The skill documents the use of 'x claude' commands to interact with various AI model providers (e.g., DeepSeek, SiliconFlow). These commands are standard for the tool's stated purpose of enhancing Claude Code's capabilities and managing sessions.- [INDIRECT_PROMPT_INJECTION]: The skill defines an interface for connecting to multiple LLM backends. While this involves processing user prompts and model responses, the documentation does not suggest any unsafe interpolation of untrusted data into privileged command contexts.
Audit Metadata