Ataru — AI skill for Claude Code
High-performance AI memory retrieval for local agent history — a Rust search core (SDK / API / JSON CLI) plus a desktop GUI.
How to install Ataru
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open lovstudio/lovcode and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Ataru does
High-performance AI memory retrieval for local agent history — a Rust search core (SDK / API / JSON CLI) plus a desktop GUI. Tantivy + Jieba keyword search, optional semantic recall, stable Turn/Run/Session/Project IDs.
Alternatives in AI
- Hindsight — State-of-the-art long-term memory for AI agents by Vectorize 6.7k ★
- Deepreasoning — A high-performance LLM inference API and Chat UI that integrates DeepSeek R1's CoT reasoning traces with Anthr 5.4k ★
- Vllm Mlx — High-performance OpenAI and Anthropic compatible LLM inference server for Apple Silicon 1.5k ★
README
Ataru
高性能 AI 记忆检索:本地会话的检索内核,以及一个桌面 GUI。
High-performance AI memory retrieval — a local search core, plus a desktop app.
macOS · Windows · Linux
为什么是 Ataru · 产品路径 · 快速开始 · Agent Skill · 架构 · 开发 · 附录与 FAQ
Why Ataru
AI 编程过程中,真正有价值的答案经常已经出现过:一次故障排查、一条关键命令、一个架构取舍,或者一段几周前才讨论过的上下文。但这些内容分散在不同 Agent 的本地会话文件里,靠记忆、目录名和人工翻找很难重新定位。
Ataru 把这些已经发生过的对话变成本地可检索的记忆层:
- 从 Claude、Codex 等本地会话来源建立统一索引。
- 用关键词、字段、短语或自然语言描述找回相关上下文。
- 在
Turn、Run、Session、Project四种粒度之间切换。 - 从命中片段回到原始会话、消息和行号,而不是只展示一段截断摘要。
- 默认离线运行;语义检索是可选增强,未配置或超时会清晰降级到关键词检索。
Ataru 的目标不是管理正在运行的 Agent,而是让过去的工作重新变得可用。
What it does
| 能力 | 解决的问题 |
|---|---|
| 统一采集 | Claude CLI、Claude App/Web、Codex 等来源的会话格式不同,Ataru 在适配层归一化它们。 |
| 增量索引 | 新消息写入后自动追赶索引,不需要每次从头扫描全部历史。 |
| 中文友好全文检索 | Tantivy + Jieba 同时覆盖中文、代码、域名、包名和错误串。 |
| 混合召回 | 关键词适合精确匹配,自然语言问题可在语义索引可用时使用混合召回。 |
| 层级聚合 | 同一命中可以按 Turn、Run、Session 或 Project 汇总,减少重复结果。 |
| 上下文回读 | 保留稳定的 Project/Session/Run/Turn 标识、片段、角色、时间和行号。 |
| Agent 接入 | GUI、CLI 和 Agent Skill 共用同一套搜索契约,不把检索逻辑复制到各个客户端。 |
Product path
Ataru 的主路径只有两步:把本地会话整理成可检索的记忆,再从命中结果回到原始上下文。
搜索召回
Related Skills
Ethan Agent
A lightweight, high-performance personal AI agent. Fast async core, pluggable tools, and open integration with
Obsidian Vault RAG Assistant
Obsidian Vault RAG Knowledge Assistant: local MiniLM embeddings + numpy retrieval, measured eval harness (reca
Landing Page Builder
A Claude Code (AI agent) skill that builds one fast, high-converting, SEO-optimized landing page — keyword-map
Maludb Core
A memory DBMS for long-term institutional memory, human-AI knowledge sharing, and contextual recall. PostgreSQ
Late CLI
High-performance AI agent for long-horizon tasks and local models. Built on empirical research. 200k+ tokens o
Karpathy LLM Wiki Stack
A comprehensive, build-ready reference for constructing a high-performance personal knowledge system using Obs
Related Agents
Mnemos Recall
Recall the most relevant Mnemos context before substantial work. Use proactively at task start to surface dura
Osf Analyze
Codebase structural analysis using codebase-retrieval. Traces dependencies, blast radius, call chains, and imp
Retro Reviewer
Reviews one prior session through ax's normalized cross-harness Turn view and emits a structured retro (tried