Skip to content

Quickstart

Complete this guide to make one local repository available to AI agents through Frigg’s shared MCP service.

  • macOS or GNU/glibc Linux for the shell installer, or another path from Installation.
  • A Git repository to index.
  • An MCP client that supports streamable HTTP.
Terminal window
curl -fsSL https://raw.githubusercontent.com/bnomei/frigg/main/scripts/install.sh \
| FRIGG_VERSION=0.9.0 sh
frigg --version

Expected version output:

frigg 0.9.0

Run from the repository root:

Terminal window
frigg init
frigg index

In non-interactive output, successful commands end with summaries like:

ok init: complete status=ok repos=1
ok index: complete status=ok mode=full
Terminal window
frigg serve

Frigg listens at:

http://127.0.0.1:37444/mcp

Keep this process running while clients use Frigg.

Preview and apply managed agent instructions plus project MCP configuration:

Terminal window
frigg adopt --target agents-md --target mcp-project --dry-run
frigg adopt --target agents-md --target mcp-project

The managed MCP entry points to the loopback HTTP service above.

Reload the client, call workspace, and adopt the repository explicitly if Frigg does not select it automatically:

{
"path": "/absolute/path/to/repo"
}

Then use search_hybrid for discovery, search_text for an exact string, and read_match or read_file for source proof.

  • ../frigg/README.md
  • ../frigg/scripts/install.sh
  • ../frigg/crates/cli/src/cli_args.rs
  • ../frigg/crates/cli/src/http_runtime.rs