Fabryk

Knowledge fabric for AI workflows.

Fabryk is a knowledge management system designed for AI workflows, providing:

• Partitioned knowledge storage with access control
• Semantic and keyword search
• Tag-based organization
• HTTP API and client libraries
• MCP (Model Context Protocol) integration

Architecture

Fabryk is composed of several crates:

• fabryk-core: Core types, traits, and errors
• fabryk-acl: Access control implementation
• fabryk-storage: Storage backend implementations
• fabryk-query: Query engine (keyword, semantic, hybrid)
• fabryk-api: HTTP API server
• fabryk-client: Rust client library
• fabryk-mcp: Model Context Protocol server
• fabryk-cli: Command-line administration tool

Quick Start

use fabryk::client::FabrykClient;

# async fn example() -> Result<(), Box<dyn std::error::Error>> {
// Create a client
let client = FabrykClient::new()?;

// TODO: Add usage examples when implementation is complete
# Ok(())
# }

Feature Flags

This crate re-exports the main Fabryk components. For CLI and MCP server binaries, see the fabryk-cli and fabryk-mcp crates respectively.