[KongchangAI]
Product

Agent Memory

Agent Memory is a key technique for addressing the stateless nature of large language models. It uses external storage mechanisms to persist an agent's working context, decision history, and learned patterns, and supports shared read/write access and conflict resolution across multiple agents.

Core Facts

Timeline (last 90 days)

Nov 14

Anthropic calls the compression process in Agent Memory that turns raw conversation streams into usable context "dreaming."

Unverified50%
Nov 14

Current engineering practice generally adopts a "frozen parameters + external memory read/write" architecture to manage Agent memory, rather than updating weights at inference time.

Unverified50%
Nov 10

For developers, practicable approaches include: writing key facts into structured notes rather than relying on model memory; setting clear read and write rules for the agent to limit what it can modify; and actively verifying whether memory has been loaded correctly when switching sessions.

Unverified50%
Nov 10

Memory loading is automatic, and users do not need to manually remind the AI what was discussed last time.

Unverified50%
Nov 10

The system can set rules for the agent, requiring it to fill in missing necessary notes or maintain a work record in a prescribed format, which is called "automatic read control" and "note rules."

Unverified50%
Nov 10

Automatic loading means the model must make trade-offs within limited context, and the quality of the retrieval strategy directly determines the accuracy of memory.

Unverified50%
Nov 10

A memory system should not be merely a passive storage-and-retrieval pipeline, but can actively construct associative structures among memories and endow memories with flexible expressive capability.

Unverified70%
Oct 30

The Agent's memory mechanism solves the stateless problem by persisting key information in external files; during operation the Agent writes valuable feedback, preferences, and lessons into the memory file, and reads it first on the next startup.

Unverified50%
Oct 30

Well-designed memory requires filtering rules, retaining only feedback with universal value for future tasks and avoiding sedimenting one-off changes into permanent rules, otherwise the memory will bloat and produce conflicts.

Unverified50%
Oct 20

The Agent Memory information officially disclosed by Databricks is limited, and details such as the storage granularity of memory, the recall mechanism, how it works with vector retrieval components, and the boundaries of memory sharing across agents have not yet been made public.

Expired50%

7 more timeline events

All Facts (20)

Verified

An Agent's memory mechanism is typically divided into two categories: short-term memory and long-term memory, with long-term memory requiring external persistent storage such as vector databases, relational databases, and graph databases.

80%
Unverified

Agent Memory 的定位是给 AI 编码代理补上长期记忆的本地记忆层,作为所有 Agent 共用的 Memory Server

95%
Unverified

Agent Memory 项目自带一个网页端的实时 Viewer,可以查看记忆的写入、搜索和回放过程

90%
Unverified

Agent Memory 提供插件、Hooks、MCP 协议以及 REST API 等多种接入方式

90%
Unverified

Agent Memory 的记忆数据存储在本地,不依赖云端服务

90%
Unverified

Agent Memory 是一个开源项目,目前已在 GitHub 上获得 10.8k stars

85%
Unverified

Agent Memory 具备语义级别的检索能力,通过向量嵌入将文本转化为高维空间中的数值向量进行语义匹配

85%
Unverified

Agent Memory 的 Readme 中给出了 Long-term Memory Retrieval Benchmark 的数据来展示其语义检索能力

85%
Unverified

在Agent Memory领域社区讨论中,强化学习与知识内化方向获得最多支持,被认为是Memory的终极目标

75%
Unverified

A memory system should not be merely a passive storage-and-retrieval pipeline, but can actively construct associative structures among memories and endow memories with flexible expressive capability.

70%
Unverified

几乎所有Agent Memory框架都围绕CRUD四个操作展开,通过embedding做向量检索,方法高度雷同

70%
Unverified

Agent Memory 支持 Claude Code、Codex、Cursor、Gemini CLI 等多种 AI 编码工具接入同一套记忆系统

60%
Unverified

Current engineering practice generally adopts a "frozen parameters + external memory read/write" architecture to manage Agent memory, rather than updating weights at inference time.

50%
Unverified

Anthropic calls the compression process in Agent Memory that turns raw conversation streams into usable context "dreaming."

50%
Unverified

For developers, practicable approaches include: writing key facts into structured notes rather than relying on model memory; setting clear read and write rules for the agent to limit what it can modify; and actively verifying whether memory has been loaded correctly when switching sessions.

50%
Unverified

Memory loading is automatic, and users do not need to manually remind the AI what was discussed last time.

50%
Unverified

Automatic loading means the model must make trade-offs within limited context, and the quality of the retrieval strategy directly determines the accuracy of memory.

50%
Unverified

The system can set rules for the agent, requiring it to fill in missing necessary notes or maintain a work record in a prescribed format, which is called "automatic read control" and "note rules."

50%
Unverified

Well-designed memory requires filtering rules, retaining only feedback with universal value for future tasks and avoiding sedimenting one-off changes into permanent rules, otherwise the memory will bloat and produce conflicts.

50%
Unverified

The Agent's memory mechanism solves the stateless problem by persisting key information in external files; during operation the Agent writes valuable feedback, preferences, and lessons into the memory file, and reads it first on the next startup.

50%

Source Articles