← Back to Skills Marketplace
Fluid Memory
by
JiagHong Phang
· GitHub ↗
· v1.0.9
9100
Downloads
5
Stars
0
Active Installs
11
Versions
Install in OpenClaw
/install fluid-memory
Description
基于艾宾浩斯遗忘曲线和访问频率的衰减模型设计的遗忘和归档机制，完全依赖openclaw原生记忆系统的拟人化流体记忆系统
Usage Guidance
Install only if you are comfortable with a local plaintext memory system. Before enabling the hook, understand that it records sent messages to ~/.openclaw/workspace/database/conversation_log.txt with no built-in retention limit or purge shown in the inspected code. Use disk encryption or strict local permissions, avoid sending secrets while it is enabled, and manually review or delete the database files if you need to clear retained content.
Capability Assessment
Credentials
Plaintext storage under ~/.openclaw/workspace/database is coherent for a local memory skill, but unbounded raw conversation logging is broader than necessary for summarized memory retention and can expose sensitive chats to local users, tools, or backups.
Install Mechanism
Installation is disclosed as a skill plus an optional copied hook, with Python package dependencies chromadb and pyyaml; those dependencies are not pinned to exact versions or hashes.
Instruction Scope
The skill documents memory triggers and local storage, but the hook behavior is broad, automatic once enabled, and the repeated description of the transcript as temporary is not matched by code that deletes or rotates it.
Persistence & Privilege
The skill persists a ChromaDB memory store, summary_buffer.json, and conversation_log.txt; the daemon is long-running if launched, and maintenance performs irreversible deletion inside the skill's memory store without a confirmation or recovery path.
Purpose & Capability
The memory, recall, archive, and decay functions fit the stated fluid-memory purpose, but the hook captures every sent message verbatim into a local transcript and maintenance can permanently delete archived ChromaDB memories after 120 days.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install fluid-memory - After installation, invoke the skill by name or use
/fluid-memory - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.9
feat: 更新描述+添加隐私声明
v1.0.8
fix: 修正HOOK.md描述+移除SKILL.md过时功能
v1.0.7
feat: 精简代码+重写简介，强调依赖原生flush触发
v1.0.6
feat: 5轮触发+Session恢复检查+完善互补说明
v1.0.5
feat: 优化静默模式 + 完善依赖声明
v1.0.4
feat: 动态查找Python路径，兼容不同环境
v1.0.3
feat: maintenance添加120天硬删除逻辑
v1.0.2
docs: 更新SKILL.md添加遗忘机制
v1.0.1
docs: 添加遗忘机制说明
v1.0.0
首发版本
v0.1.0
fluid-memory v0.1.0
- 初始发布：基于流体认知架构的记忆系统。
- 支持「植入记忆」「唤起记忆」「遗忘」和「状态查询」四大工具接口。
- 记忆会随时间/使用强化或遗忘，实现更自然、拟人化的记忆管理。
- 工具调用示例与最佳实践说明，便于理解与集成。
Metadata
Frequently Asked Questions
What is Fluid Memory?
基于艾宾浩斯遗忘曲线和访问频率的衰减模型设计的遗忘和归档机制，完全依赖openclaw原生记忆系统的拟人化流体记忆系统. It is an AI Agent Skill for Claude Code / OpenClaw, with 9100 downloads so far.
How do I install Fluid Memory?
Run "/install fluid-memory" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Fluid Memory free?
Yes, Fluid Memory is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Fluid Memory support?
Fluid Memory is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Fluid Memory?
It is built and maintained by JiagHong Phang (@againta); the current version is v1.0.9.
More Skills