The context layer for AI-native engineering teams.
LLMs got good. The workflows around them did not.
Cruxet is building the missing layer between AI agents and production engineering, so teams can ship with AI the same way they ship with humans: with version control, security boundaries, observability, and clear ownership.
We are starting with the connective tissue: MCP.
@cruxet/mcp-audit - Local, zero-setup security linter for MCP client configs.
npx @cruxet/mcp-audit finds command injection, hardcoded secrets, insecure transports, and known CVEs across Cursor, Claude, Windsurf, VSCode, Continue, Codex, and Zed. No account, no API calls, no data leaves your machine.
A platform for orchestrating MCP-driven workflows across teams - context as a first-class artifact, with the same lifecycle discipline you already apply to code.
If that resonates, follow the org or open an issue on mcp-audit - we read every one.
Built in Istanbul.