Metadata-Version: 2.4
Name: langgraph-checkpoint-hyperspace
Version: 4.2.0
Summary: Standalone LangGraph CheckpointSaver & Cross-Thread Store backed by HyperspaceDB CSR KAG Graph and Lorentz Hyperbolic Manifolds
Author-email: YARlabs <hi@yar.ink>
License: Apache-2.0
Project-URL: Homepage, https://yar.ink
Project-URL: Repository, https://github.com/YARlabs/langgraph-checkpoint-hyperspace
Project-URL: Documentation, https://yar.ink/docs
Keywords: hyperspacedb,langgraph,checkpoint,ai-agents,knowledge-graph,lorentz-manifold
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: hyperspacedb>=4.2.0
Provides-Extra: langgraph
Requires-Dist: langgraph-checkpoint>=2.0.0; extra == "langgraph"
Provides-Extra: dev
Requires-Dist: pytest>=8.0.0; extra == "dev"
Requires-Dist: pytest-asyncio>=0.23.0; extra == "dev"
Dynamic: license-file

# `langgraph-checkpoint-hyperspace` — Standalone LangGraph Checkpointer & Store for HyperspaceDB

[![PyPI](https://img.shields.io/pypi/v/langgraph-checkpoint-hyperspace.svg)](https://pypi.org/project/langgraph-checkpoint-hyperspace/)
[![License: Apache-2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](./LICENSE)

A lightweight, zero-monolith **LangGraph CheckpointSaver** and **Cross-Thread Memory Store** backed by **[HyperspaceDB](https://yar.ink)** CSR KAG Topology and Lorentz Hyperbolic Manifolds.

Unlike `langchain-hyperspace`, this standalone package has **zero dependency on `langchain` or `langchain-core`**, making it ideal for lean production LangGraph deployments.

---

## Key Capabilities

1. **CSR KAG Execution Lineage & Fork Tracking**
   - Automatically wires `STEP_TRANSITION` edges between consecutive checkpoints and `FORKED_FROM` edges when branching or time-traveling from an earlier state.
2. **Lorentz Hyperboloid State Embeddings**
   - Embeds serialized state snapshots onto the upper sheet of the $d$-dimensional Lorentz hyperboloid ($\mathbb{H}^{d-1}$) where radial distance encodes execution step depth.
3. **Semantic State Recall & Time-Travel**
   - `saver.search_checkpoints("query")` recalls relevant past states across execution branches using Lorentzian geodesic similarity.
   - `saver.traverse_lineage(thread_id, start_checkpoint_id)` walks the KAG execution graph in $O(1)$ per hop.

---

## Installation

```bash
pip install langgraph-checkpoint-hyperspace
```

---

## Quick Start

```python
from langgraph_checkpoint_hyperspace import HyperspaceCheckpointSaver

checkpointer = HyperspaceCheckpointSaver(
    host="localhost",
    port=50051,
    collection_name="agent_checkpoints",
)

# Pass directly to any LangGraph StateGraph.compile(checkpointer=checkpointer)
```

---

## License

Apache-2.0. Copyright © 2026 **YARlabs**.

