Metadata-Version: 2.5
Name: sql-evidence-graph
Version: 0.1.1
Summary: SQL-first evidence graph on SQLite: documents, entities, relations, and one-hop expansion. Ranking is supplied by the caller.
Project-URL: Homepage, https://github.com/kroq86/sql-evidence-graph
Project-URL: Repository, https://github.com/kroq86/sql-evidence-graph
Project-URL: Issues, https://github.com/kroq86/sql-evidence-graph/issues
Author-email: KirillOstapenko <djkroq@gmail.com>
License-Expression: MIT
License-File: LICENSE
Keywords: evidence,graph,retrieval,sqlite
Classifier: Development Status :: 3 - Alpha
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Python :: 3.14
Requires-Python: >=3.11
Description-Content-Type: text/markdown

# sql-evidence-graph

SQLite evidence graph. One hop from seed documents along shared entities. The caller supplies scores. This package does not embed text, stem words, or call a model.

```python
from sql_evidence_graph import EvidenceGraph, order_by_score

graph = EvidenceGraph("knowledge.db")
graph.create()
graph.link("doc-1", "feature:payments", "payments", "feature", "body", "payments settle overnight")
neighbors = graph.expand(["doc-1"])
ordered = order_by_score(neighbors, {"doc-2": 1.5})
```

`expand` returns every other document that shares an entity. It does not read the question and does not drop a neighbor. `order_by_score` puts positive caller scores first.
