--- slug: "iflytekopensource-adapters" source_type: "readme" source_url: "https://cdn.jsdelivr.net/gh/iflytek/memflywheel@main/README.md" repo: "https://github.com/iflytek/memflywheel" source_file: "README.md" branch: "main" --- # MemFlywheel

MemFlywheel
Turn every Agent run into a smarter start for the next one!

English | 简体中文

![MemFlywheel overview](https://github.com/iflytek/memflywheel/raw/HEAD/docs/assets/readme/01-overview.png) MemFlywheel adds a file-native memory flywheel to Agent Harnesses: recall before execution, extract after execution, and evolve repeated workflows into learned skills.

File-native Markdown memories, source traces, and learned skills stay inspectable and diffable.	Progressive recall Pre-recall plus layered reads from index cues to evidence.
Post-run learning Turn-end extraction and dream consolidation keep memory moving.	Harness-native Pi, Hermes, OpenCode, and OpenClaw are supported through npm packages.

## Why It Exists Give your Agent a memory flywheel: recall before it acts, learn after it runs, and understand you better each time. The host Agent Harness owns lifecycle, model access, auth, and tools; MemFlywheel owns the memory and learning loop. ## How It Works ```text Agent Harness | | lifecycle / model / auth / tools v MemFlywheel | |-- pre-recall -> MEMORY.md index cues |-- progressive read -> memory bodies -> source traces -> learned skills |-- turn-end -> durable memory extraction |-- idle -> dream consolidation and repair `-- repeated work -> reusable learned skills ```


Memory lifecycle Recall, extract, consolidate, and keep evidence close to the file-native store.	Skill flywheel Repeated work evolves into reusable learned skills the Agent can inspect and reuse.

## Quick Start Pi: ```sh pi install npm:@iflytekopensource/adapters ``` Hermes: ```sh npm install -g @iflytekopensource/hermes memflywheel-hermes-install hermes config set memory.provider memflywheel ``` OpenCode: ```sh opencode plugin @iflytekopensource/adapters --global opencode run --dir /path/to/project "your task" ``` OpenClaw: ```sh openclaw plugins install npm:@iflytekopensource/adapters openclaw config set plugins.slots.memory memflywheel openclaw config set plugins.entries.memflywheel.hooks.allowConversationAccess true openclaw config set plugins.entries.memflywheel.hooks.allowPromptInjection true openclaw gateway run --force ``` MemFlywheel installs into each host as a native memory plugin. The host keeps owning models, tools, permissions, and sessions; MemFlywheel adds recall, turn-end extraction, dream consolidation, and learned skills. Embedding pre-recall is optional. Without it, MemFlywheel still works and injects up to 200 generated `MEMORY.md` index lines directly. Once your memory index grows beyond that, start any OpenAI-compatible embeddings endpoint and export these variables before starting the host; pre-recall then turns on automatically and injects only the most relevant index entries. ```sh export MEMFLYWHEEL_EMBEDDING_ENDPOINT="https://embedding-gateway.example.com/v1" export MEMFLYWHEEL_EMBEDDING_API_KEY="..." export MEMFLYWHEEL_EMBEDDING_MODEL="text-embedding-3-small" ``` Host setup, embedding pre-recall, verification, and troubleshooting live in [`docs/integrations.md`](https://github.com/iflytek/memflywheel/blob/HEAD/docs/integrations.md). ## Install Packages | Package | Role | | ------------------------------------------------------------------------------------------ | -------------------------------------------------------------------------- | | [`@iflytekopensource/adapters`](https://www.npmjs.com/package/@iflytekopensource/adapters) | Pi, OpenCode, OpenClaw, and the shared host-adapter runtime used by Hermes | | [`@iflytekopensource/hermes`](https://www.npmjs.com/package/@iflytekopensource/hermes) | Hermes MemoryProvider installer and skill mirror | Internal workspace packages keep the code split by responsibility; users install only the host package they need. ## Evaluation MemFlywheel uses LoCoMo-oriented regression checks to keep long-term-memory behavior measurable while the recall, extraction, and learned-skill loops evolve. See [`docs/evaluation.md`](https://github.com/iflytek/memflywheel/blob/HEAD/docs/evaluation.md). ## Documentation | Document | Content | | ------------------------------------------------------------------ | --------------------------------------------------------------------------------- | | [`docs/architecture.md`](https://github.com/iflytek/memflywheel/blob/HEAD/docs/architecture.md) | Storage layout, recall, extraction, dream, skill loop, package boundaries | | [`docs/integrations.md`](https://github.com/iflytek/memflywheel/blob/HEAD/docs/integrations.md) | Pi, Hermes, OpenCode, OpenClaw, embedding pre-recall, SDK hooks, adapter boundary | | [`docs/evaluation.md`](https://github.com/iflytek/memflywheel/blob/HEAD/docs/evaluation.md) | LoCoMo position and local regression checks | | [`docs/release.md`](https://github.com/iflytek/memflywheel/blob/HEAD/docs/release.md) | Versioning, npm release channel, publish checklist | | [`CHANGELOG.md`](https://github.com/iflytek/memflywheel/blob/HEAD/CHANGELOG.md) | Release notes for public npm package versions | | [`NOTICE`](https://github.com/iflytek/memflywheel/tree/HEAD/NOTICE), [`THIRD_PARTY_LICENSES`](https://github.com/iflytek/memflywheel/tree/HEAD/THIRD_PARTY_LICENSES) | Project notice and third-party license disclosure | ## Open-Source Boundary MemFlywheel is a foundation component inside an Agent Harness. It stays file-native, model-agnostic, and host-first; it does not absorb the main Agent, model service, tool permissions, or skill execution into itself. ## 💬 Community Join the Astron Open Source Community (WeCom Group) to discuss and collaborate: