openarangodb

name	openarangodb
description	Agent memory and enterprise features for ArangoDB Community Edition. Use when storing, searching, or managing agent memories in ArangoDB, or when the user asks about CDC, audit logging, vector search, graph traversals, encryption validation, or backup for ArangoDB.

OpenArangoDB Skill

Enterprise-equivalent features for ArangoDB Community Edition — CDC, audit, encryption, agent memory, vector search, graph traversals, and more.

When to Use

• Storing or retrieving agent memories
• Setting up ArangoDB with enterprise features (audit, CDC, encryption)
• Vector similarity search on stored memories
• Multi-layer retrieval (exact, tag, semantic, temporal) with RRF fusion
• Graph traversals and SmartGraph-like partitioning
• Change data capture and audit logging
• Backup, replication, or encryption validation

Installation

# Library only
pip install OpenArangoDB

# With MCP server support
pip install "OpenArangoDB[mcp]"

# All optional features
pip install "OpenArangoDB[all]"

MCP Server Configuration

Add to Claude Code settings (~/.claude/settings.json):

{
  "mcpServers": {
    "openarangodb": {
      "command": "openarangodb",
      "args": ["serve"],
      "env": {
        "ARANGODB_HOST": "http://localhost:8529",
        "ARANGODB_DATABASE": "argondb",
        "ARANGODB_USERNAME": "root",
        "ARANGODB_PASSWORD": ""
      }
    }
  }
}

Or with Docker:

{
  "mcpServers": {
    "openarangodb": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "openarangodb:latest"],
      "env": {
        "ARANGODB_HOST": "http://host.docker.internal:8529"
      }
    }
  }
}

Available MCP Tools

Tool	Description
memory_insert	Insert a new memory (content, tags, entity, category, agent_id)
memory_get	Retrieve a memory by ID
memory_search	Vector similarity search
memory_update	Update an existing memory
memory_delete	Soft-delete a memory
memory_supersede	Replace a memory with a newer version (creates supersession chain)
retrieval_search	Multi-layer search with RRF fusion (exact + tag + semantic + temporal)
audit_query	Query audit logs (by agent, operation, time range)
changes_since	Get CDC change events since a revision
encryption_check	Check encryption at rest status

Python API Quick Reference

from open_arangodb import ArangoDB, Memory, AgentScope

# Connect
db = ArangoDB(
    host="http://localhost:8529",
    database="argondb",
    audit_enabled=True,
    cdc_enabled=True,
)

# Insert
memory = Memory(id="m1", content="The user prefers dark mode", tags=["preference"])
scope = AgentScope(agent_id="agent-1")
db.insert(memory, scope=scope)

# Search
results = db.search("dark mode", limit=5)

# Supersede (version a memory)
db.supersede("m1", Memory(id="m2", content="The user switched to light mode"))

# Get CDC changes
changes = db.get_changes(since_rev="12345")

# Multi-layer retrieval
from open_arangodb import RetrievalRequest
results = db.retrieve(RetrievalRequest(query="user preferences", entity="user-1"))

Environment Variables

Variable	Default	Description
ARANGODB_HOST	http://localhost:8529	ArangoDB URL
ARANGODB_DATABASE	argondb	Database name
ARANGODB_USERNAME	root	Username
ARANGODB_PASSWORD	(empty)	Password
ARANGODB_AUDIT	true	Enable audit logging
ARANGODB_CDC	true	Enable CDC
ARANGODB_RETRIEVAL	false	Enable multi-layer retrieval
ARANGODB_TEMPORAL	false	Enable temporal engine
ARANGODB_GRAPH	false	Enable graph features

CLI Commands

openarangodb serve          # Start MCP server (stdio)
openarangodb health         # Check ArangoDB connection
openarangodb version        # Show version
openarangodb encrypt-check  # Check encryption status