lovstudio/Ataru

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.

What it solves

Ataru provides a local, searchable memory layer for AI conversations. It solves the problem of valuable answers, commands, and architectural decisions becoming lost across fragmented local session files from various AI agents (like Claude and Codex), making it difficult to retrieve specific context from past work.

How it works

Ataru indexes local transcript files from multiple AI sources and normalizes them into a unified format. It uses a combination of keyword search (via Tantivy and Jieba) and optional semantic search (via OpenAI-compatible embeddings and SQLite) to recall information. Users can search at different granularities—Turn, Run, Session, or Project—and then jump directly back to the original session, message, and line number in the source file.

Who it’s for

Developers and power users who use multiple AI agents for coding and troubleshooting and need a way to quickly find and reuse previous AI-generated solutions and discussions.

Highlights

  • Unified Indexing: Normalizes session formats from Claude CLI, Claude App/Web, and Codex into a single searchable index.
  • Local-First Privacy: Runs offline by default; original transcripts are read-only, and semantic search is an optional opt-in feature.
  • Agent-Ready: Includes a dedicated "Search Skill" and JSON CLI, allowing other AI agents to programmatically retrieve historical context.
  • Multi-Granularity Search: Allows switching between Turn, Run, Session, and Project levels to reduce noise and find exactly the right level of context.
  • Incremental Updates: Automatically tracks changes in local session files to update the index without needing a full rescan.

Related

  • Project
    vshulcz/deja-vu

    A unified memory layer for coding agents that indexes local session history to provide cross-agent recall of previous decisions and fixes.

  • Project
    Dicklesworthstone/coding_agent_session_search

    A high-performance TUI and CLI tool that indexes and searches local history from multiple AI coding agents in a unified, searchable timeline.

  • Project
    raine/claude-history

    A terminal-based history browser for Claude Code, Pi, and Oh My Pi that enables searching, viewing, and managing local AI agent conversation transcripts.

  • Dispatch
    ctx: Local Search for Coding Agent History

    ctx is an open-source Rust-based CLI that indexes local coding agent session logs into SQLite to provide fast, token-efficient retrieval of past decisions and failed attempts.

  • Project
    ctxrs/ctx

    ctx is a Rust‑based CLI that indexes local coding‑assistant logs (Claude, Codex, Cursor, etc.) and provides fast lexical or optional semantic search over those transcripts. It lets agents retrieve exact prior messages, tool calls, and, with the paid ctx pro add‑on, map any file/line/commit/PR back to the originating agent session (a “git blame” for AI‑generated code). All processing runs locally, preserving privacy and reducing token consumption.