AnythingLLM: Prior to 1.13.0, the filesystem-search-files agent skill passes its LLM-controlled pattern…
Affects: AnythingLLM
Securaity has published 4 advisories for AnythingLLM: 1 High, 2 Medium, 1 Low.
AnythingLLM is a chat interface made by Mintplex Labs. It gives people a web screen where they can type questions to AI models and read the answers. It is often connected to models that the organization chooses, rather than to one fixed model. The people who usually install and run it are IT staff, developers or enthusiasts who want a single chat screen for the AI models they have picked. Other staff then use that screen to chat with those models.
Affects: AnythingLLM
Affects: AnythingLLM
Affects: AnythingLLM
Affects: AnythingLLM