42-evey
evey-memory-adaptive
Adaptive memory plugin for Hermes Agent that scores and decays memories, retaining high-value ones while fading low-value ones.
Repository last updated Jul 18, 2026
Overview
Evey Memory Adaptive is a plugin for the Hermes Agent ecosystem that introduces importance-based memory management. It assigns a score to each memory and applies a decay function over time, ensuring that high-value memories persist while low-value ones gradually fade away. This mimics natural forgetting and helps agents focus on relevant information.
The plugin is designed for developers building autonomous agents that need efficient long-term memory. By automatically prioritizing important memories, it reduces noise and improves decision-making without manual intervention. It integrates seamlessly with Hermes Agent via a simple plugin configuration.
Built with Python and released under the MIT license, Evey Memory Adaptive is open-source and easy to customize. It is ideal for AI agents that operate over extended periods and require a dynamic memory system that adapts to changing contexts.
Highlights
- Importance scoring for each memory entry
- Automatic decay of low-value memories over time
- High-value memories persist longer
- Seamless integration with Hermes Agent
- Open-source and MIT licensed
Features
Importance Scoring
Each memory is assigned a score based on its relevance, ensuring critical information is retained.
Decay Mechanism
Memories with low importance scores gradually decay and are removed, preventing memory bloat.
Plugin Architecture
Easy to install and configure as a plugin within the Hermes Agent ecosystem.
Installation
Commands are reproduced exactly as published by the project maintainers. Always check the upstream repository for the current instructions.
git clone https://github.com/42-evey/hermes-plugins.git
cp -r hermes-plugins/evey-* ~/.hermes/plugins/
cp hermes-plugins/evey_utils.py ~/.hermes/plugins/
# Restart hermes-agentRequirements
- Hermes Agent installed
- Python 3.8+
How it's built
The plugin is structured as a standard Hermes Agent plugin with a plugin.yaml manifest and a Python package evey-memory-adaptive. The core logic resides in __init__.py, which implements the scoring and decay algorithms. It integrates via the plugin interface, requiring no modifications to the agent core.
Memory persistence is handled internally, and the plugin can be configured through the plugin.yaml file. The architecture is minimal, focusing on the adaptive memory lifecycle: score, decay, and prune.
Source
This summary was written for Neura Market from the project's own repository. Install commands, identifiers and licence details are reproduced unchanged. The 42-evey/evey-memory-adaptive repository is the authoritative source and the place to file issues or contribute.
Category: Memory & Knowledge · Licensed under MIT