Is user data kept private?

Yes, the architecture is user-scoped, meaning memory is isolated per user/actor with granular privacy controls and management.

What is the difference between episodic and short-term memory in this skill?

Short-term memory stores raw turn-by-turn session context for immediate use, while episodic memory consists of long-term facts extracted from conversations via a reflection engine for use in future sessions.

Can I manually manage the memory records?

Absolutely. While the skill supports automatic extraction, it also provides operations to manually create, update, or delete specific memory records.

How fast is memory retrieval?

The skill is optimized for high performance, targeting a memory retrieval latency of less than 100ms.

Bedrock AgentCore Memory

Name: Bedrock AgentCore Memory
Author: adaptationio

byadaptationio

数据科学与机器学习

Integrates persistent episodic and short-term memory into Amazon Bedrock agents to maintain user context and learn from interactions across sessions.

关于

Bedrock AgentCore Memory provides a robust framework for building intelligent agents that remember user preferences and improve over time. By combining short-term session context with long-term episodic memory extracted through a background reflection engine, this skill allows Claude to retrieve relevant historical facts with sub-100ms latency. It is ideal for developers creating personalized experiences, technical support assistants, or long-running workflows where maintaining a continuous 'thread of thought' across multiple user interactions is critical.

主要功能

Short-term session context management with sliding window strategies
Episodic memory extraction for long-term fact retention
0 GitHub stars
Automatic background reflection and deduplication of facts
User-scoped memory isolation for privacy and security
Semantic retrieval for context-aware memory lookups

使用场景

Creating technical support agents that track past issues and successful solutions
Building personalized travel or shopping assistants that remember user preferences
Maintaining context for complex, multi-day project management workflows

关于

主要功能

Short-term session context management with sliding window strategies
Episodic memory extraction for long-term fact retention
0 GitHub stars
Automatic background reflection and deduplication of facts
User-scoped memory isolation for privacy and security
Semantic retrieval for context-aware memory lookups

使用场景

Creating technical support agents that track past issues and successful solutions
Building personalized travel or shopping assistants that remember user preferences
Maintaining context for complex, multi-day project management workflows