sglang-setup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly asks the agent to check for an HF_TOKEN and to pass it inline (e.g., -e HF_TOKEN="..."/-e HF_TOKEN="" in the docker run), requiring the LLM to include a secret verbatim in generated commands, which is high-risk secret handling.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill pulls and runs the external Docker image lmsysorg/sglang:latest-cu130 at runtime (docker pull / docker run), which executes remote code from that image and is a required dependency, so it can directly control agent behavior.