An independent Ebbinghaus forgetting curve module providing standardized memory decay calculations and spaced repetition scheduling for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install forgetting-curve
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 forgetting-curve 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 Forgetting Curve?
The Forgetting Curve skill is a robust implementation of the Ebbinghaus forgetting curve, designed to help developers manage information retention within the Openclaw Skills ecosystem. It provides standardized calculations for memory decay and strength, allowing AI agents to simulate cognitive processes by prioritizing knowledge based on its age and relevance. By integrating this module, you can move beyond static data retrieval and create dynamic, human-like memory systems that recognize when information needs to be refreshed or phased out.
Forgetting Curve Use Cases
- Simulating human-like memory decay in AI personal assistants to prioritize recent or important information.
- Implementing Spaced Repetition Systems (SRS) for AI-driven language learning or educational platforms.
- Optimizing RAG (Retrieval-Augmented Generation) pipelines by weighting retrieved documents based on their calculated memory strength.
- Managing long-term memory in autonomous agents by identifying and pruning stale data through standardized decay models.
How Forgetting Curve Works
- Initialize the module with specific parameters such as half-life (defaulting to 30 days) and initial memory strength.
- Calculate the decay factor by comparing the current timestamp against the last interaction date (age in days).
- Apply the decay function (Exponential or Power Law) to the original memory strength to determine its current value.
- Utilize the Spaced Repetition Scheduler to determine the optimal next review interval based on current strength and historical success factors.
- Batch process large memory sets to update the global state of the agent's knowledge base efficiently.
Forgetting Curve Setup
To install the module as part of your local environment, use pip:
pip install -e .
Alternatively, for developers utilizing the Openclaw Skills registry via ClawHub, you can install it directly:
clawhub install forgetting-curve
Forgetting Curve Data Schema & Taxonomy
The Forgetting Curve skill manages memory metadata through a structured set of configurable parameters:
| Parameter | Type | Default | Description |
|---|---|---|---|
| half_life_days | float | 30.0 | The time required for memory strength to reduce by 50%. |
| initial_strength | float | 1.0 | The starting strength value (0.0 to 1.0) assigned to new memories. |
| base_interval | float | 1.0 | The baseline interval in days for the first repetition cycle. |
| strength_factor | float | 1.5 | The multiplier used to calculate increasing intervals in the SRS cycle. |
| minimum_strength | float | 0.1 | The threshold below which a memory is considered forgotten. |
Forgetting Curve Advanced Features
- Support for multiple decay models including Ebbinghaus exponential decay and Power Law decay.
- Customizable decay functions for domain-specific memory behavior (e.g., shorter half-lives for temporary task data).
- Multi-tier memory system support, allowing separate curves for Short-Term Memory (STM) and Long-Term Memory (LTM).
- High-performance batch processing capable of calculating 10,000 decay states in approximately 5ms.
- Integration hooks for popular frameworks like CortexGraph and Memory Sync Enhanced.
SKILL.md
Loading
Related Openclaw Skills
Brand Visual Generator
A comprehensive design system engine that establishes typography, color palettes, and spatial systems for consistent brand presentation across web and print.
Docs Logger
A cloud-native logging skill that automatically organizes and appends daily activity logs to Google Docs via the PortEden CLI.
Travel Deal Finder
An AI-powered travel agent that automates real-time price comparisons and credit card point optimization for global travel.
iFind Financial Data MCP
A professional financial data interface for querying A-shares, public funds, macroeconomic indicators, and news via the Model Context Protocol.
Ad Creative AI Strategy & Generation
An expert performance creative skill for generating, iterating, and scaling platform-specific ad copy and visuals.
Apple Media Control
An integrated macOS automation tool for discovering and managing Apple media devices, HomePods, and AirPlay speakers via pyatv and Airfoil.
Featured*
