AI Session Search: Ultra-Fast AI Agent Session Analysis

Introduction

AI Session Search (aise) is an ultra-fast, Rust-powered tool designed for comprehensive searching and analysis of your local AI agent coding sessions. It addresses the common challenge of disparate session formats by unifying data from various AI tools into a single, searchable index. Whether you use Claude Code, Claude Desktop, ChatGPT Codex, Cursor, Antigravity, Pi, Google AI Studio, or Gemini CLI, aise provides a seamless way to manage and retrieve your AI-generated history.

Find the project on GitHub.

Why Use and Key Benefits

aise offers a robust solution for developers working with multiple AI coding assistants, providing several key advantages:

• Unified Session Management: It parses and indexes nine different session formats, allowing you to search across all your AI interactions from a single interface. This eliminates the need to navigate separate logs and tools.
• Blazing Fast Performance: Built with Rust, aise is engineered for speed. It can search millions of messages in under a second, ensuring quick access to critical information without performance bottlenecks.
• Versatile Access Options: Access your indexed data through multiple interfaces: a powerful Command Line Interface (CLI), an MCP (Model Context Protocol) server for real-time agent queries, a Rust library for direct integration, and a Python API for scripting and automation.
• Enhanced Session Recovery: Easily recover files an agent wrote or edited, view the complete history of a file across sessions, and even jump back into a past session to resume work.
• Deep Agent Behavior Analysis: Gain insights into agent performance by searching for specific events like "harness notices" to understand why an agent stopped, looped, or was blocked. You can also analyze subagent runs and identify corrections made to agent outputs.
• Broad AI Tool Compatibility: aise supports a wide array of popular AI coding agents and platforms, ensuring your entire local AI coding history is covered.
• Efficient Incremental Indexing: The indexing process is incremental, meaning only changed sessions are re-parsed during updates, significantly reducing refresh times and maintaining high performance.

Installation

Installing aise is straightforward. You can use uv (recommended) or Rust's cargo.

Using uv (recommended):

uv tool install ai-session-search && aise integrations install

Using cargo:

cargo install ai-session-search --locked --version '^1.0.0-rc' && aise integrations install

After installation, run aise reindex to build your initial session index.

Examples

Here are some common tasks and example commands to get you started with aise:

• Index your sessions:

  aise reindex
  

• List what you've been working on:

  aise list
  aise search "database migration" --provider codex
  

• Find a specific moment and read around it:

  aise messages search "permission denied" --role user
  aise show SESSION_ID
  aise resume SESSION_ID
  

• Recover a file:

  aise files history src/app.py
  aise files extract src/app.py > src/app.py
  

For more detailed usage, aise COMMAND --help provides comprehensive documentation for every parameter.

Links

• GitHub Repository: https://github.com/ahundt/ai-session-search
• PyPI: https://pypi.org/project/ai-session-search/
• Crates.io: https://crates.io/crates/ai-session-search