ui-ux-pro-max
Fail
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: HIGHPRIVILEGE_ESCALATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The
SKILL.mdfile provides instructions for the agent to execute commands with elevated privileges, specificallysudo apt update && sudo apt install python3for installing dependencies on Linux. - [COMMAND_EXECUTION]: The skill's operation involves executing various shell commands, including environment checks (
python3 --version), package manager commands (brew,winget), and a local search utility (scripts/search.py). - [INDIRECT_PROMPT_INJECTION]: The skill implements a search function that takes user keywords and injects matched content from local CSV files into the agent's context, creating a vulnerability surface for indirect prompt injection.
- Ingestion points: User-provided search keywords are passed to
scripts/search.pyas described inSKILL.md. - Boundary markers: The search results are formatted with headers but lack explicit safety delimiters or instructions to ignore embedded commands.
- Capability inventory: The skill environment allows for shell command and script execution.
- Sanitization: The Python search implementation (
scripts/core.py) does not include sanitization or filtering to prevent malicious instructions in user queries or the database from influencing the agent.
Recommendations
- AI detected serious security threats
Audit Metadata