moorcheh-client 0.1.5

Self-hosted, open-source Moorcheh - The Information-Theoretic Search Engine for RAG & Agentic Memory

PyPI · Documentation · Contributing · Apache 2.0

Install

pip install moorcheh-client

Requirements: Python 3.10+, Docker, and an embedding provider (Ollama, OpenAI, or Cohere).

Quick start

moorcheh up          # first run prompts for embedding/LLM config
moorcheh status      # health + quota

API: http://localhost:8080 · Data: ~/.moorcheh/data

from moorcheh import MoorchehClient

client = MoorchehClient("http://localhost:8080")

client.namespaces.create("docs", type="text")
job = client.documents.upload("docs", documents=[
    {"id": "doc-1", "text": "Hello from Moorcheh on-prem"},
])
# poll job["job_id"] until completed, then search or answer

results = client.similarity_search.query(
    namespaces=["docs"],
    query="Hello",
    top_k=5,
)

answer = client.answer.generate(
    namespace="docs",
    query="What is in my documents?",
)

moorcheh down        # stops containers; data is kept

CLI

Common commands: moorcheh up, moorcheh down, moorcheh status, moorcheh namespace-create, moorcheh upload-documents, moorcheh search, moorcheh answer.

Run moorcheh --help or see docs.moorcheh.ai/on-prem for the full CLI and API reference.

Test endpoints

With the server running:

python test.py

Project layout

Contributing

See CONTRIBUTING.md.

License

Apache License 2.0