{"repo":"unbody-io/adapt","free":true,"listed":false,"github":"https://github.com/unbody-io/adapt","clone":"git clone https://github.com/unbody-io/adapt.git","description":"A self-evolving memory layer for LLMs.","language":"TypeScript","stars":33,"topics":["ai-agents","ai-memory","ai-tools","knowledge-base","npm-package","typescript"],"license":"MIT","category":"ai-agents","readme_excerpt":"Adapt A memory layer that learns. Instead of storing and retrieving, Adapt observes incoming data, builds understanding, and reshapes its own structure over time. It answers questions that databases and RAG pipelines can't — the ones that require paying attention as data flows in. Documentation · Changelog · Releases · npm · Issues --- Install Adapt uses the Vercel AI SDK for LLM access. Install a provider: Quick start Features - Self-evolving — creates, merges, splits, and removes Neurons based on usage - Any LLM — AI SDK by default, BYO runtime via the AdaptLLMPlugin contract - Pluggable stores — in-memory or SQLite - Modular — use the Brain orchestrator or TextNeuron / ListNeuron standalone - Persistent — Brain.restore(path) rehydrates from SQLite; same for standalone neurons - Runs anywhere — Node, Bun, and Electron (ESM + CJS builds) Limitations - Requires models with structured output and tool calling support - Local model support (Ollama, LMStudio) not fully tested yet - Not a database — builds understanding, doesn't store raw data - Experimental ( 0.0.x ) — expect breaking changes Contributing Found a bug or have an idea? Open an issue. PRs welcome. --- MIT — Unbody","default_branch":null,"files":null,"tree":[],"storefront":"/r/unbody-io","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/unbody-io/adapt/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}