Blarify Quickstart¶

Status: Shipped. All commands on this page exist in the current binary. Cross-check with amplihack index-code --help and amplihack query-code --help.

This how-to imports a blarify JSON export into the code graph and runs a sanity query.

Before you start¶

You need:

amplihack-rs built (cargo build --release) and on PATH
A blarify JSON export for your project (produced by upstream blarify)
Optional: SCIP indexers if you also want to use index-scip (see Index a Project)

Steps¶

1. Locate or generate the blarify JSON¶

Blarify itself is out of scope for this guide; produce the JSON with your existing blarify pipeline. The importer expects the path on the command line.

2. Import the JSON into the code graph¶

amplihack index-code path/to/blarify-export.json

To put the resulting database somewhere other than the default location:

amplihack index-code path/to/blarify-export.json --db-path ./custom.lbug

The legacy --kuzu-path alias is still accepted for backward compatibility but is hidden in --help.

3. Verify the import¶

amplihack query-code stats

You should see non-zero counts for Files, Classes, and Functions. If everything is zero, the JSON was empty or its schema did not match what the importer expects — re-run blarify and confirm its output is valid before re-importing.

4. Run a structural query¶

amplihack query-code search --name MyClass
amplihack query-code callers --name run_tui

For the full subcommand list:

amplihack query-code --help

When to prefer `index-scip` instead¶

If your language has a SCIP indexer available (scip-python, scip-typescript, rust-analyzer --scip, …), prefer amplihack index-scip — it produces the same downstream graph but with richer cross-reference information. See Index a Project.