The Vanna library uses a prompt function to present the user with visualized results, it is possible to alter the prompt using prompt injection and run arbitrary Python code instead of the intended visualization code. Specifically - allowing external input to the library’s “ask” method with "visualize" set to True (default behavior) leads to remote code execution.
CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H
This vulnerability has a 15.18% probability of being exploited in the next 30 days, ranking higher than 97% of all scored CVEs.
View this score breakdown or calculate a custom score
Learn how severity scores are calculated and what they mean
Best practices for deciding which vulnerabilities to address first
Essential guide to Common Vulnerabilities and Exposures
Understand how CVEs relate to underlying weakness types