AgentVectorDB (AVDB)
The Cognitive Core for Your AI Agents - Powered by LanceDB
📚 Superagentic AI Introduction
📚 Project Documentation
🌟 Overview
AgentVectorDB (AVDB) is a specialized memory management system developed by Superagentic AI. Built on top of LanceDB's powerful vector database capabilities, it provides optimized cognitive architecture for AI agents.
🤝 Built with LanceDB
We extend LanceDB's robust foundation with agent-specific features:
- Agent memory patterns
- Importance scoring
- Context management
- Cognitive state handling
✨ Key Features
Core Capabilities
- 📝 Persistent Storage: File-based, no server required
- 🔍 Semantic Search: Efficient ANN search with filtering
- ⚡ Async Support: High-performance async/await API
- 🎯 Agent-Optimized: Purpose-built for AI systems
Advanced Features
- 🔄 Memory Lifecycle: Complete CRUD operations
- 📊 Batch Processing: Efficient bulk operations
- 🧹 Smart Pruning: Intelligent memory management
- 🔧 Flexible Schema: Dynamic Pydantic schemas
- ⏱️ Time Tracking: Automatic timestamps
📦 Installation
# Basic installation
pip install agentvectordb
# Development installation
git clone https://github.com/superagenticai/agentvectordb.git
cd agentvectordb
pip install -e ".[dev]"
#For windows if you encounter UnicodeDecodeError set $env:PYTHONUTF8=1
🚀 Quick Start
from agentvectordb import AgentVectorDBStore
from agentvectordb.embeddings import DefaultTextEmbeddingFunction
# Initialize store
store = AgentVectorDBStore(db_path="./agent_db")
ef = DefaultTextEmbeddingFunction(dimension=384)
# Create collection
memories = store.get_or_create_collection(
name="agent_memories",
embedding_function=ef
)
# Add memories (minimum 8 recommended)
initial_memories = [
{
"content": "User prefers dark mode",
"type": "preference",
"importance_score": 0.8
},
# Add more memories...
]
# Add batch
memories.add_batch(initial_memories)
# Query memories
results = memories.query(
query_text="user preferences",
k=2
)
🛠️ API Overview
Store Classes
AgentVectorDBStore
store = AgentVectorDBStore(db_path="./db")
Methods:
get_or_create_collection()list_collections()
AsyncAgentVectorDBStore
store = AsyncAgentVectorDBStore(db_path="./db")
Collection Classes
AgentMemoryCollection
Methods:
add(): Add single memoryadd_batch(): Add multiple memoriesquery(): Search memoriesprune_memories(): Remove old memoriesdelete(): Remove memoriescount(): Get collection size
📚 Advanced Usage
Custom Embedding Functions
from agentvectordb.embeddings import BaseEmbeddingFunction
class CustomEmbedder(BaseEmbeddingFunction):
def __init__(self, dimension=384):
super().__init__(dimension=dimension)
def embed(self, texts):
# Your embedding logic here
return vectors
Memory Management
# Prune old memories
pruned = collection.prune_memories(
max_age_seconds=7*24*3600, # 7 days
min_importance_score=0.3
)
# Complex queries
results = collection.query(
query_text="important task",
filter_sql="type = 'task' AND importance_score > 0.8",
k=5
)
Async Operations
async def handle_memories():
store = AsyncAgentVectorDBStore(db_path="./async_db")
collection = await store.get_or_create_collection(
name="async_memories",
embedding_function=ef
)
await collection.add_batch(memories)
results = await collection.query(
query_text="search term",
k=5
)
🎯 Use Cases
- Personal AI assistants
- Customer service bots
- Research agents
- Task automation agents
- Knowledge management systems
- Learning systems
🔄 Memory Types
- Episodic memories
- Semantic knowledge
- Procedural information
- Short-term observations
- Long-term knowledge
🛣️ Roadmap
Upcoming features:
- Enhanced filter builders
- Reflection/summarization helpers
- Schema evolution support
- Memory consolidation
- Extended embedding support
- Performance optimizations
🤝 Contributing
We welcome contributions! See our Contributing Guide.
📄 License
Licensed under Apache 2.0 - same as LanceDB. See LICENSE.
🙋♂️ Support
- GitHub Issues: Technical support
- Email: support@super-agentic.ai
- Enterprise: enterprise@super-agentic.ai
🙏 Acknowledgments
Built with ❤️ by Superagentic AI using LanceDB
AgentVectorDB: Empower Your AI Agents with Memory!
Metadata
Release files for agentvectordb 0.0.5
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| agentvectordb-0.0.5.tar.gz | 32.8 kB | Without provenance | Details for agentvectordb-0.0.5.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| agentvectordb-0.0.5-py3-none-any.whl | Python 3 | none | any | Details for agentvectordb-0.0.5-py3-none-any.whl |
Total release size: 70.5 kB
Release files / agentvectordb-0.0.5.tar.gz
| Download URL | agentvectordb-0.0.5.tar.gz |
|---|---|
| Size | 32.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
56ec451db09b12bca472a33b42c05a72c528f87a40d2dd9b9a1e02213b4150a2
|
|
BLAKE2b-256 checksum How to use checksums |
5c20a947b42e82c7960fc7b95ad377d981abbe7e39f4601a3324331295e9ca09
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.12.13 {"installer":{"name":"uv","version":"0.12.13","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Release files / agentvectordb-0.0.5-py3-none-any.whl
| Download URL | agentvectordb-0.0.5-py3-none-any.whl |
|---|---|
| Size | 37.8 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
eed32851a0acffaeaadb80262ef8d8574b69622435a0ac4b837f9bd8241b33e1
|
|
BLAKE2b-256 checksum How to use checksums |
349fb0ef01ed5f8d3a26caa2595838b15cf63dcb5555fcda0e600c8b55f2d211
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.12.13 {"installer":{"name":"uv","version":"0.12.13","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Security for agentvectordb 0.0.5
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.