KGLite

An embedded Cypher dialect for LLM-agent workloads. A knowledge graph that runs inside your process — load data, query with Cypher, and hand the graph to an agent via the bundled MCP server. The embedded path needs no database service; one .kgl file can move between Python and Rust bindings.

The engine is a pure-Rust crate (kglite); the wheel (pip install kglite) is a PyO3 wrapper around it. Bolt and MCP protocol servers are standalone Rust binaries that wrap the same engine. The .kgl file format is portable across all bindings.

Start here

  1. Install kglite (Python) or add the kglite crate (Rust).

  2. Build a graph with inline records, DataFrames, Cypher, or a companion project such as codingest/kglite-datasets.

  3. Query with Cypher or the fluent API; use Session/Transaction when a failed mutation must roll back.

  4. Save a .kgl, or serve it through the CLI, MCP, or Bolt binary.

Python quickstart · Cypher reference · Fluent API · Rust quickstart · Operators and deployment · Reference · 0.13 → 0.14 migration

Cypher first

Cypher is the primary query surface — agents already know it, and the engine targets an explicitly documented openCypher-compatible subset (including three-valued NULL logic), checked with independently authored local contracts and optional Neo4j differential runs. DataFrame loaders add_nodes() / add_connections() exist to get bulk data in; once it’s in, you query with Cypher.

Embedded, in-process

No database service; import and go

LLM-agent surface

Bundled MCP server + describe() schema for system prompts

Cypher subset, honest semantics

Querying + mutations + text_score() for semantic search

In-memory by default

Mapped + disk modes for Wikidata-scale; in-memory is the design centre

Label model

One primary type + optional secondary labels — see multi-label rationale

One-file persistence

.kgl snapshots — copy, share, reload elsewhere

Rust-embeddable

Pure-Rust core; embed without PyO3 — see Rust track

Ecosystem

kglite is the engine. Two companion projects build graphs it serves — each released and versioned on its own cadence:

  • kglite — the embedded Cypher knowledge-graph engine (this project): graph + Cypher + fluent API + bundled MCP server.

  • codingest — parses codebases into code graphs (14 languages, web-framework route detection). Build with it, query the .kgl here. Requires kglite ≥ 0.14.

  • kglite-datasets — fetch-build-cache loaders for public registries (SEC EDGAR, Wikidata, Sodir).

  • sonagram — turns a local music library into a kglite knowledge graph via sonara audio analysis (tempo, energy, mood, key); AI agents curate playlists over it through a simple bundled skill and CLI (pip install sonagram).

Coming from 0.13? The code-graph builder and dataset loaders moved out of the wheel in 0.14 — see the 0.13 → 0.14 migration guide. Pin back anytime with pip install "kglite<0.14".

Pick your track

  • Python guide — pip install kglite, then import kglite. The headline track; covers data loading, Cypher, the MCP server, agents.

  • Rust guide — embed the engine in a Rust binary (cargo add kglite). For graph-as-a-library use cases without the Python wheel.

  • Operators — choose and run the CLI, MCP, or Bolt binary; storage, auth/TLS, and deployment guidance.

  • Reference — Python, Cypher, fluent, Rust, C ABI, and CLI reference surfaces.

  • Concepts — architecture + design decisions + contributor docs.