daedra

Daedra is a high-performance MCP server written in Rust that provides web search and page fetching capabilities. It exposes tools to perform DuckDuckGo-based web searches and to fetch and extract content from web pages, returning results in convenient formats. You can run Daedra as an MCP server using STDIO transport (suitable for integrations with Claude Desktop and other MCP clients) or via HTTP SSE transport for a lightweight server with streaming responses. The server can be used either as an IPC service through STDIO or as a standalone HTTP service, and it also provides a Rust library for deeper integration. After starting the MCP server, you can send requests to search the web or fetch pages through the MCP protocol, enabling AI assistants to obtain search results or page content without leaving the current workflow.

Prerequisites:

• Rust and Cargo installed (Rust 1.91 or later)
• Internet access to fetch dependencies

Install from crates.io (binary):

cargo install daedra

Install from source:

git clone https://github.com/dirmacs/daedra.git
cd daedra
cargo install --path .

Using Cargo in a Rust project (as a library):

[dependencies]
daedra = "0.1"

Quick start (as an MCP server):

# Start the server via STDIO transport (default for MCP clients like Claude Desktop)
daedra serve --transport stdio --quiet

Or start via SSE/HTTP transport:

daedra serve --transport sse --port 3000 --host 127.0.0.1

Notes and tips:

• The CLI supports two transports: stdio (for direct MCP clients) and SSE (HTTP) for web-based integrations. Use --quiet to minimize logs to avoid interfering with JSON-RPC streams.
• Environment variable: RUST_LOG can be set to control log verbosity (e.g., RUST_LOG=debug).
• The server offers web search (DuckDuckGo) and page fetch capabilities, with results that can be cached for performance.
• If running via the CLI in a production environment, consider configuring port/host for SSE transport and enabling caching with --cache-ttl to suit your workload.
• For advanced usage, you can leverage the Rust library directly in your application to perform searches and fetch content programmatically.