free-claude-code-ai-desktop-app

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). The GitHub release is suspicious because it comes from a non-official/impersonating-looking account and distributes a direct ZIP (with an executable that the instructions tell you to run as Administrator), which is a high-risk distribution vector; the other URLs (localhost:11434, ara.so, and api.anthropic.com) are low-risk/legitimate endpoints.

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

• Potentially malicious external URL detected (high risk: 1.00). The Windows installation step downloads and extracts an executable from GitHub ("https://github.com/anthropic-claude-code-ai/free-claude-code-ai-desktop-app/releases/download/claude-code/free-claude-app.zip") and then starts ClaudeCodeAI.exe, which clearly fetches and executes remote code at runtime.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill explicitly instructs installing into "C:\Program Files..." and to "Run the app as Administrator" (including Start-Process -Verb RunAs) which encourages obtaining elevated privileges and modifying system-level files/locations, so it pushes actions that can change the machine state and require admin access.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)