migrate-from-daytona

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's SKILL.md and referenced SDK docs explicitly show cloning arbitrary public repositories into sandboxes (e.g., sandbox.process.exec / git clone "https://github.com/user/repo.git /app"), which causes the agent to fetch and read/run untrusted, user-generated web content that can materially influence subsequent commands and behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill fetches MCP tool definitions at runtime (via bl_tools / the sandbox MCP endpoint https://<SANDBOX_URL>/mcp), which returns tool definitions used by AI frameworks and therefore directly controls agent behavior/instructions.