Eli Ainhorn

**Name of the Vulnerable Software and Affected Versions** Context7 versions prior to 2.1.2 **Description** A prompt injection issue exists where unsanitized content can be injected through the Custom AI Instructions feature served via the MCP server. This allows attackers to execute malicious instructions in connected AI coding agents. Potential impacts include the exfiltration of credentials from environment files to an external service and the destructive deletion of files on the victim's machine during routine library documentation requests. **Recommendations** Update Context7 to version 2.1.2 or later. Restrict the use of the Custom AI Instructions feature until the update is applied.

**Name of the Vulnerable Software and Affected Versions** Kong Konnect MCP server versions prior to 1.0.0 **Description** An issue exists where the server fails to properly validate content returned to the Large Language Model (LLM). This allows a remote attacker to perform an indirect prompt injection by placing malicious text within data that the AI agent reads. Consequently, the agent may interpret these smuggled instructions as its own and execute unintended API requests against Kong Konnect using its own permissions, which could lead to the exposure of sensitive data. **Recommendations** Update Kong Konnect MCP server to version 1.0.0 or later.