VantaDB × CrewAI

CrewAI Tool adapter for VantaDB.

Install

Not on PyPI yet. The vantadb-crewai package builds and passes twine check locally; it will go live with the first adapters-v* tag release. Until then, install from source.

# Today, from a repo checkout
cd integrations/crewai && pip install .

Install from PyPI (after first release)

pip install vantadb-crewai

Quickstart

from crewai import Agent, Task, Crew
from vantadb_crewai import VantaDBTool

rag_tool = VantaDBTool(
    name="Memory Search",
    description="Search stored documents in VantaDB",
    db_path="./my_data",
    namespace="docs",
)

agent = Agent(
    role="Assistant",
    goal="Answer questions using stored knowledge",
    tools=[rag_tool],
)

API

• VantaDBTool(name, description, db_path, namespace) — CrewAI-compatible RAG tool
• VantaDBMemoryBackend(db_path, namespace) — StorageBackend for unified Memory (CrewAI ≥1.14): save/search/delete/update/get_record/list_records/ get_scope_info/list_scopes/list_categories/count/reset + async asave/asearch/adelete

from crewai import Memory
from vantadb_crewai import VantaDBMemoryBackend

memory = Memory(storage=VantaDBMemoryBackend(db_path="./my_data"))
memory.remember("We decided to use PostgreSQL.", scope="/project/decisions")
print(memory.recall("What database did we choose?"))

System fields live under reserved __mem_* metadata keys; your metadata is untouched. Records without embeddings are invisible to vector search (same as LanceDB).

Why VantaDB?

• Embedded & local-first: the storage engine is a Rust library embedded in your process — no server to deploy, no network hop; data lives in your filesystem.
• Persistent hybrid search: vectors + BM25 text search out of the box, where CrewAI's native memory covers only short-term session recall.
• Zero-setup alternative to hosted stacks: unlike Zep (requires a server) or Cognee (spins up its own knowledge-graph runtime), VantaDB is a plain library you import.

Development

pip install -e .