Wrste / pheromone-network

# pheromone_network
**Memory that strengthens with use and fades with disuse.**
A zero-dependency, embeddable local association and recall kernel: frequently used relationships grow stronger, while idle relationships naturally decay.
English | 中文
`TypeScript` · `Node >= 18` · `0 runtime deps` · `offline-friendly`
## What problem does it solve?
Many systems need to answer the same question: **given a new input, which pieces of past information are worth using again?**
`pheromone_network` packages that loop into a small local component. You provide vectors; it learns local associations, scores candidates, and lets unused relationships fade over time. It does not require a remote model, a vector database, a training dataset, a network connection, or a GPU.
Use it as a standalone recall layer, or place it in front of search, caching, recommendation, rules, or agent memory.
## Install from npm
The package is publicly available on npm as version `0.1.0`. It includes compiled CommonJS output and TypeScript declarations, so applications can install and import it directly:
 [code] 
 [code] 
Package page: 
## Browser demo
Try the live demo directly in your browser: **Open the online Memory Lab**. It demonstrates memory observation, ranked recall, repeated reinforcement, frequency weighting, and time based pheromone decay. You can also open `examples/index.html` locally. See `examples/README.md` for local serving and GitHub Pages instructions.
## Why use it?
- **Adapts through real usage**: repeated hits reinforce the relationships that matter in your product.
- **Lets stale context fade**: evaporation and time based decay keep old relationships from dominating forever.
- **Small and inspectable**: local connections, explicit weights, and cosine scores are easy to inspect.
- **Simple deployment**: zero runtime dependencies, no network requirement, and no GPU requirement.
- **Works with existing vectors**: use model embeddings, hashed features, or the built in `ngramEmbed` encoder.
## Where can it be used?
| Scenario | How to use it | Typical outcome |
| --- | --- | --- |
| **Semantic caching** | Associate request vectors with previous requests; reinforce reused entries and decay old ones | Fewer repeated computations for local APIs, tools, and generated content |
| **Recommendation and personalization** | Write behavior or content features; reinforce clicks, saves, and repeated actions | Lightweight interest ranking and personalized shortcuts |
| **Document, ticket, and knowledge deduplication** | Score a new item against historical vectors before review | Similar tickets, duplicate FAQs, and existing solutions surface early |
| **Log and event recall** | Encode event features or text n-grams and retrieve similar historical events | Faster incident investigation and runbook matching |
| **Offline and edge applications** | Keep associations on the device or inside a tenant or project scope | Desktop tools, browser extensions, IoT, and privacy sensitive syste