Deep Agents Memory & Filesystem

Name: Deep Agents Memory & Filesystem
Availability: InStock
Author: LangChain

Provide agents with ephemeral, persistent, and filesystem storage via pluggable backends.

1.1k|86|Updated Jan 22, 2026

Search Tags:#backend #memory #persistence #filesystem #storage #deep agents

Ready to run

npx skills add https://github.com/langchain-ai/langchain-skills --skill deep-agents-memory-filesystem

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Deep Agents Memory & Filesystem
Source: https://github.com/langchain-ai/langchain-skills/tree/main/config/skills/deep-agents-memory
Command: npx skills add https://github.com/langchain-ai/langchain-skills --skill deep-agents-memory-filesystem

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the need for AI agents to have persistent memory, temporary storage, and direct access to the filesystem, enabling more complex and stateful interactions.

Core Features & Use Cases

Ephemeral Memory (StateBackend): Stores data within a single thread, useful for short-term context.
Persistent Memory (StoreBackend): Saves data across threads and sessions, ideal for long-term knowledge.
Filesystem Access (FilesystemBackend): Allows agents to read, write, and manipulate files on disk, with security considerations.
Hybrid Storage (CompositeBackend): Routes different file paths to different backend types for flexible storage strategies.
Filesystem Utilities: Provides ls, read_file, write_file, edit_file, glob, and grep operations.

Quick Start

Configure the agent to use a CompositeBackend that routes files in /memories/ to a persistent StoreBackend.

Frequently Asked Questions about Deep Agents Memory & Filesystem

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

FAQPage Schema

How do I add persistent memory to AI agents so they remember context across sessions?▼

Persistent memory for AI agents is achieved using a StoreBackend to save data across multiple threads and sessions. This allows agents to retain long-term knowledge for stateful interactions.

Can AI agents read and write files directly from the filesystem?▼

Yes, AI agents can read and write files directly from the filesystem using a FilesystemBackend. It provides utilities including ls, read_file, write_file, edit_file, glob, and grep for disk operations.

What is the best way to manage both temporary and persistent data in agent workflows?▼

Managing temporary and persistent data in agent workflows is best done using a CompositeBackend for hybrid storage routing. It directs different file paths to ephemeral StateBackend or persistent StoreBackend types.

How do I store temporary context for an agent within a single thread?▼

To store temporary context for an agent within a single thread, use a StateBackend for ephemeral memory. This retains short-term data active during the workflow without persisting to disk or cross-session storage.

Does this agent memory storage approach work for disk-based data management?▼

Yes, this agent memory storage approach works for disk-based data management through the FilesystemBackend. It allows agents to manipulate files on disk with necessary security considerations.