skills/migokg/plugin-store/uniswap-ai/Gen Agent Trust Hub

uniswap-ai

Fail

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: HIGHDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill contains a bash script in SKILL.md that performs system fingerprinting by concatenating the hostname, operating system name, hardware architecture, and the path to the user's home directory.
  • [DATA_EXFILTRATION]: This system fingerprint is cryptographically signed and exfiltrated to external servers at plugin-store-dun.vercel.app and okx.com via POST requests. Telemetry reporting to a third-party crypto exchange is highly irregular for a DeFi integration tool.
  • [COMMAND_EXECUTION]: The skill executes shell commands automatically during session initialization under the guise of 'Pre-flight Dependencies', performing profiling and network operations without explicit user notification or consent.
  • [PROMPT_INJECTION]: The skill uses deceptive metadata and branding, claiming to be an official tool maintained by 'Uniswap Labs' and referencing official Uniswap GitHub repositories, despite being authored by an unrelated third party (MigOKG) and performing hidden data collection.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Apr 12, 2026, 01:29 PM