AgentDB Memory Patterns

Implement persistent memory and pattern learning for AI agents using AgentDB.

Updated May 8, 2026
One-click install
npx skills add https://github.com/FrekiManagarm/d-chambaud --skill agentdb-memory-patterns-frekimanagarm
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: AgentDB Memory Patterns
Source: https://github.com/FrekiManagarm/d-chambaud/tree/main/.agents/skills/agentdb-memory-patterns
Command: npx skills add https://github.com/FrekiManagarm/d-chambaud --skill agentdb-memory-patterns-frekimanagarm

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires agentdb, agentic-flow.

What problem does it solve? AI agents lose conversation context between sessions and cannot learn from past interactions, forcing developers to rebuild state management, vector search, and memory consolidation from scratch. ## Core Features & Use Cases - Session and Long-Term Memory: Store conversation history, user preferences, and facts with vector embeddings for semantic retrieval across sessions. - Pattern Learning with 9 Algorithms: Train agents using Decision Transformer, Q-Learning, Actor-Critic, and other reinforcement learning plugins via CLI or API. - ReasoningBank Integration: Migrate legacy ReasoningBank databases and use reasoning agents for context synthesis, memory optimization, and experience curation. - Use Case: Build a customer support chatbot that remembers each user's preferences, retrieves relevant past conversations via vector search, and improves its responses by learning from successful interactions. ## Quick Start Initialize an AgentDB database and set up session memory for my chat agent so it can remember conversations across sessions.

Frequently Asked Questions about AgentDB Memory Patterns

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add persistent memory to an AI agent?▼

Initialize an AgentDB database with npx agentdb init, then use the createAgentDBAdapter API to store conversation patterns with vector embeddings. Retrieve relevant context using retrieveWithReasoning with configurable k and MMR options.

How to migrate from ReasoningBank to AgentDB?▼

Use the migrateToAgentDB function from agentic-flow/reasoningbank, passing the legacy .swarm/memory.db source path and the new .agentdb/reasoningbank.db destination. The migration reports how many patterns were transferred and maintains full API compatibility.

What learning algorithms does AgentDB support?▼

AgentDB provides nine learning plugins including Decision Transformer, Q-Learning, SARSA, Actor-Critic, Active Learning, Adversarial Training, Curriculum Learning, Federated Learning, and Multi-task Learning. Create plugins with npx agentdb create-plugin using built-in templates.

Does AgentDB work with Codex via MCP?▼

Yes, AgentDB includes an MCP server started with npx agentdb mcp. Register it once with Codex using the mcp add command, after which Codex can access the vector database tools directly.

Why is my AgentDB database growing too large?▼

Large databases result from storing unquantized embeddings. Enable binary quantization for 32x memory reduction or scalar for 4x reduction, and run memory consolidation with importance-based strategies and relevance thresholds to prune low-value patterns.