Kaval is a Model Context Protocol (MCP) server: Verify the facts an agent's action depends on before it acts — ALLOW, REVIEW, or BLOCK. In practice that means any MCP-compatible AI assistant can call Kaval's tools directly — the model decides when to use them in a conversation or agent run.
It runs locally: the npm package @usekaval/mcp speaks MCP over stdio on your machine. Configuration happens through environment variables — you'll need KAVAL_API_KEY. The current release is v0.7.1, published under the Apache-2.0 license, with source at LufeMC/kaval-clients on GitHub.
Kaval is listed under AI & Agents on mcp.site and works with any MCP client — Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, Zed and the rest of the ecosystem — using the install snippets below. If you maintain Kaval for LufeMC, claim this listing to verify ownership, earn the Verified badge, and keep the details current.