autoteam-f-chatgpt-team-rotation
Fail
Audited by Gen Agent Trust Hub on Apr 29, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill clones a software repository from an unverified third-party GitHub account (ZRainbow1275/AutoTeam-F) and fetches browser binaries via Playwright.\n- [COMMAND_EXECUTION]: Installation and operational workflows involve executing shell scripts (setup.sh) and tool-specific commands through the uv package manager, allowing for extensive interaction with the host system.\n- [REMOTE_CODE_EXECUTION]: By downloading and immediately running a setup script from a non-trusted repository, the skill creates a path for arbitrary code execution on the user's local machine.\n- [DATA_EXFILTRATION]: The tool is designed to read local account databases containing sensitive OAuth authentication tokens and synchronize them to external API endpoints such as CLIProxyAPI.\n- [PROMPT_INJECTION]: The skill processes data from external temporary email services to automate registration, which establishes a surface for indirect prompt injection where malicious email content could manipulate the agent's account management actions.
Recommendations
- AI detected serious security threats
Audit Metadata