Memoria

Secure · Auditable · Programmable Memory for AI Agents

Persistent memory layer for AI agents (Kiro, Cursor, Claude Code, any MCP-compatible agent) with Git-level version control — snapshots, branches, rollback, and full audit trail.

Full documentation: https://github.com/matrixorigin/Memoria

Two Modes

	Managed / Remote	Self-hosted
Flag	`--api-url` + `--token`	`--db-url`
Requires	Nothing — connect to existing server	MatrixOne DB + embedding config
When	Team / SaaS, admin gives you a URL + token	Personal setup, local dev

Install

# Managed / remote mode — no extras needed
pip install memoria

# Self-hosted embedded mode — choose an embedding provider:
pip install "memoria[openai-embedding]"   # OpenAI / SiliconFlow / any OpenAI-compatible endpoint
pip install "memoria[local-embedding]"    # Local sentence-transformers (~900MB download)

# If no NVIDIA GPU available, install CPU-only PyTorch first to avoid large CUDA dependencies:
pip install torch --index-url https://download.pytorch.org/whl/cpu
pip install "memoria[local-embedding]"

Quick Start

Managed mode (no database, no embedding setup)

If your team or provider gives you a server URL and API token:

cd your-project
memoria init --api-url "https://your-server:8100" --token "sk-your-key..."

Restart your AI tool — done.

Self-hosted mode (run your own database)

# 1. Start MatrixOne
git clone https://github.com/matrixorigin/Memoria.git && cd Memoria
docker compose up -d

# 2. Configure
cd your-project
memoria init --db-url "mysql+pymysql://root:111@localhost:6001/memoria"

# With OpenAI-compatible embedding (recommended over local model)
memoria init --db-url "mysql+pymysql://root:111@localhost:6001/memoria" \
             --embedding-provider openai \
             --embedding-base-url https://api.siliconflow.cn/v1 \
             --embedding-api-key sk-... \
             --embedding-model BAAI/bge-m3 \
             --embedding-dim 1024

memoria init auto-detects Kiro / Cursor / Claude and writes MCP config + steering rules.

Verify

memoria status

Embedding Providers (self-hosted mode only)

Provider	Quality	Privacy	Cost	First-use latency
Local (default)	Good	✅ Data never leaves machine	Free	~900MB download on first use
OpenAI / SiliconFlow	Better	⚠️ Text sent to API	API key required	None
Custom service	Varies	Depends on host	Self-hosted	None

Managed mode users don't need to configure embedding — the server handles it.

memory_store, memory_retrieve, memory_correct, memory_purge, memory_search, memory_profile, memory_snapshot, memory_rollback, memory_branch, memory_merge, memory_diff, and more.

License

Apache-2.0 © MatrixOrigin

Metadata

Release files for mo-memoria 0.1.24

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Built distribution (wheel)

Table of built distributions (wheels) for mo-memoria 0.1.24
File	Interpreter	ABI	Platform	Reset
mo_memoria-0.1.24-py3-none-any.whl 336.1 kB Mar 16, 2026 Without provenance	Python 3	none	any	Details

Release files / mo_memoria-0.1.24-py3-none-any.whl

Download URL	mo_memoria-0.1.24-py3-none-any.whl
Size	336.1 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`3c1edd274b274c22d24afe3adbcc63cc1f1bd63c28b2c4f55ec811ccc14d3891`
BLAKE2b-256 checksum How to use checksums	`1b4c0189aa4d7a33d70afd14e05b74b8a4a4b10dc4feb04d6b7425a36ed46bfe`
Upload date	Mar 16, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.11.14`

Security for mo-memoria 0.1.24

Report project as malware

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.