A sophisticated memory management system for AI agents that prioritizes high-value context while pruning low-value noise.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install remember
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install remember using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Remember?
The Remember skill is designed to solve the problem of information overload in AI memory systems. Instead of simple logging, this skill curates persistent memory by filtering what matters, organizing data by function, and decaying information that is no longer relevant. By implementing this within the Openclaw Skills ecosystem, developers can ensure their agents maintain a high-signal environment for better decision-making.
This framework moves beyond basic retrieval by categorizing information into specific high-value silos like commitments, preferences, and corrections. It ensures that the agent doesn't just remember everything, but specifically remembers the right thing at the right time, making it an essential component for any advanced AI workflow.
Remember Use Cases
- Tracking explicit user preferences and stylistic boundaries to ensure output consistency.
- Managing long-term project commitments and deadlines to act as a reliable assistant.
- Recording learned corrections and past mistakes to prevent repetitive errors.
- Maintaining organizational context and relationship roles across multiple sessions.
How Remember Works
- Analyze incoming interaction data to identify high-value information versus transient noise.
- Categorize the validated information into functional files such as commitments, preferences, or decisions.
- Append metadata to each entry, including the recording date, source hint, and confidence level.
- Execute a staleness test to prune or archive outdated data every 30 days or upon context completion.
- Handle contradictions by updating the active record while maintaining a log of the change for transparency.
Remember Setup
To initialize the memory framework for Openclaw Skills, set up the following directory structure in your agent's workspace:
mkdir -p memory/contexts
touch memory/commitments.md memory/preferences.md memory/corrections.md memory/decisions.md memory/relationships.md
Ensure your agent has read/write permissions to this directory to allow for real-time memory updates and hygiene tasks.
Remember Data Schema & Taxonomy
The skill utilizes a functional file-based taxonomy to organize data for optimal retrieval:
| File | Purpose | Data Type |
|---|---|---|
commitments.md |
Promises, deadlines, and deliverables | Task/Date |
preferences.md |
User likes, dislikes, and style constraints | Preference |
corrections.md |
Negative feedback and mistakes to avoid | Constraint |
decisions.md |
Rationale for specific paths chosen | Log |
contexts/*.md |
State data for specific projects or clients | State |
Remember Advanced Features
- Functional organization over content-based organization for faster contextual retrieval.
- Automated memory hygiene protocols to prune inactive contexts and completed commitments.
- Conflict resolution that logs historical updates (
Was: X, Now: Y) rather than silent overwriting. - Granular user control commands including explicit 'forget' and 'transparency' queries.
- Confidence-weighted memory entries to distinguish between explicit user statements and agent inferences.
SKILL.md
Loading
Related Openclaw Skills
Regex Expert
A comprehensive guide and utility for writing efficient, secure, and cross-engine compatible regular expressions.
Self Reflection
An AI skill that enables agents to self-critique, learn from mistakes, and proactively surface lessons to reduce revision rounds.
Redis
A high-performance skill for managing Redis caching, message queues, and complex data structures with atomicity and memory safety.
Recruiter
A comprehensive AI assistant for sourcing candidates, screening resumes, and managing professional hiring pipelines with a focus on impact and equity.
Remote Desktop
A comprehensive remote access management skill for establishing secure RDP, VNC, and SSH X11 connections with automated tunneling.
Render Deploy
Streamline your application deployment process on Render with automated codebase analysis and Blueprint generation.
Featured*
