computer-use

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). High-risk: the code auto-installs and executes third-party "skills" (npx commands) and invokes subprocess.run(..., shell=True), creating an untrusted supply-chain / remote-code-execution vector even if not explicitly a backdoor.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runner invokes external installers at runtime (subprocess.run) such as "npx skills add web-infra-dev/midscene-skills@computer-automation" and "npx skills add am-will/codex-skills@gemini-computer-use", which will fetch and execute remote code and thus directly control executed behavior.