Claude Memory Pro is an evolved unified memory system that integrates a 5-layer architecture with intelligent token budgeting and cost optimization for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install claude-memory-pro
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 claude-memory-pro 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 Claude Memory Pro?
Claude Memory Pro represents a significant upgrade in how AI agents handle persistence and resources. By distilling the core architecture of Claude Code, this Openclaw Skills contribution provides a sophisticated framework for cross-session learning and personal knowledge base construction. It moves beyond simple context windows by categorizing information into distinct layers, ensuring that the most relevant data is always available without bloating the context.
This skill is particularly focused on efficiency and sustainability. It doesn't just store information; it manages the financial and computational cost of AI interactions. With built-in token budget management and marginal utility detection, it prevents runaway API costs while ensuring the agent continues to evolve based on user feedback and project evolution.
Claude Memory Pro Use Cases
- Managing long-term user preferences and working styles across multiple sessions.
- Tracking complex project contexts and technical decisions to ensure consistency.
- Building a reusable repository of procedures and entity definitions for recurring tasks.
- Optimizing AI operational costs through intelligent token tracking and cache utilization.
- Facilitating cross-project learning where the agent reflects on past interactions to improve future performance.
How Claude Memory Pro Works
- Context Loading: At the start of a session, the system reads the central index to load relevant long-term and project memories.
- Real-time Monitoring: During the conversation, 'Hot Memory' tracks the current turn while the budget manager monitors token usage and costs.
- Budgetary Decisioning: The system evaluates token consumption against thresholds (e.g., 75%, 90%) to decide whether to continue or nudge the user for completion.
- Marginal Utility Check: It detects diminishing returns in long conversations to prevent wasted tokens on repetitive outputs.
- Post-Session Reflection: After the interaction, the system evaluates which information is worth keeping and saves it to the appropriate memory layer (User, Feedback, Project, or Reference).
- Index Update: The central memory index is updated to ensure the new knowledge is discoverable in the next session.
Claude Memory Pro Setup
To initialize the advanced 5-layer memory structure for your agent using Openclaw Skills, run the following command in your project root:
mkdir -p memory/{hot,session,typed/{user,feedback,project,reference},bank/{entities,procedures},reflections}
Ensure your agent configuration points to this directory as its primary persistence layer.
Claude Memory Pro Data Schema & Taxonomy
Claude Memory Pro organizes data into a hierarchical structure to balance retrieval speed and storage depth:
| Layer | Directory | Purpose |
|---|---|---|
| 1: Hot | memory/hot/ |
Current turn context and immediate transient data. |
| 2: Session | memory/session/ |
Historical records of specific interaction sessions. |
| 3: Typed | memory/typed/ |
Categorized long-term memory (User, Feedback, Project, Reference). |
| 4: Bank | memory/bank/ |
Structured entity archives and reusable procedural workflows. |
| 5: Reflection | memory/reflections/ |
Meta-analysis and self-improvement logs for the agent. |
Claude Memory Pro Advanced Features
- Intelligent Token Budgeting: Configurable thresholds for completion, warning, and automatic continuation limits.
- Marginal Utility Detection: Automatically stops processing if the delta in information gain drops below a specific token threshold.
- Multi-Dimensional Cost Tracking: Detailed monitoring of input/output tokens, cache hits, and external search costs.
- Automatic Cache Optimization: Strategies to maximize cache_read hits and compress context to reduce input costs.
- Evolutionary Reflection: A dedicated layer for the agent to store self-corrections and behavioral improvements based on user feedback.
SKILL.md
Loading
Related Openclaw Skills
ORBCAFE Kanban + Detail
A specialized skill for building interactive Kanban boards with drag-and-drop workflows and integrated detail navigation using the ORBCAFE UI ecosystem.
AI Security Guard Pro
A comprehensive security framework for AI coding agents that provides real-time command scanning, sandbox isolation, and hierarchical permission management.
Cost Optimizer Pro
An intelligent AI cost optimization system for tracking token budgets, detecting marginal returns, and generating detailed expense reports.
ORBCAFE Pivot + AINav
An advanced UI toolkit for building interactive, hook-first pivot tables and AI-powered voice navigation workflows.
Invoice Generator
A professional billing API for generating GST-compliant invoices and automating email distribution to clients.
Resume Builder
An automated document generation API that transforms structured data into professional, ATS-friendly resumes and tailored cover letters.
Featured*
