ahundt

ai-session-search: Search Local AI Coding Sessions

ai-session-search indexes local transcripts from multiple AI coding tools and lets you search sessions, messages, and edited files. Use it from a Rust-powered CLI, MCP server, Rust library, or Python API.

RustApache-2.0AICLIMCP
View on GitHub Compare
Stars
45
Forks
1
Stars, last 30 days
+33

Our take

Health score

39/100

How it is scored
Activity
71
Community
21
Issues
21
Pull requests
43

+33 stars in the last 30 days

Project history

Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.

Selected snapshot

Stars

79

Forks

1

Change since Nov 16, 2026: +4 stars·0 forks

Overview

ai-session-search (aise) turns locally stored transcripts from supported AI coding tools into a searchable index. It addresses a common gap: useful decisions, explanations, and generated file versions are scattered across different providers and transcript formats.

The same Rust core serves a command-line tool, an MCP server, a Rust library, and a Python API. It is intended for developers who want to retrieve or analyze their own local AI coding history, rather than search conversations stored only in cloud accounts.

Key Features

  • Indexes local sessions from nine listed sources, including Claude Code, Codex, Cursor, Gemini CLI, and Pi.
  • Searches sessions and individual messages, with filters for provider, path, date, role, and other indexed fields.
  • Finds and reconstructs file versions recorded in sessions.
  • Exposes search to compatible coding agents through an MCP server.
  • Provides Rust and Python APIs backed by the same core services as the CLI.
  • Supports exports and corpus analysis, alongside an interactive terminal browser.
  • Refreshes the index incrementally as local session files change.

Use Cases

  • A developer can find the earlier session where an agent explained or fixed a recurring error.
  • Someone comparing AI coding tools can search across their local transcripts instead of learning each tool's storage format.
  • A developer can recover or inspect a file version an agent created during a previous task.
  • A coding agent can query the user's indexed history through MCP while helping with a current task.
  • A team or individual building local tooling can integrate session search through the Rust crate or Python package.

Getting Started

Install with uv and set up detected integrations:

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

Then run aise reindex to index discovered local sessions. See the README for provider details, configuration, and other installation options.

Alternatives

  • ai-memory: ai-memory turns coding-agent activity into a git-backed Markdown wiki for handoffs and retrieval, rather than providing dedicated transcript and edited-file search.
ProjectLanguageLicenseStarsStatus
ai-session-searchRustApache-2.045Not checked yet
ai-memoryRustMIT9.1kNew

Considerations

  • It reads local session files. Cloud-only conversations are not retrieved from provider accounts.
  • The README says no stable version is published yet; the documented install commands use a release candidate.
  • Integrations are a separate, explicit setup step. Installing the package alone does not configure MCP clients or scan transcripts.
  • Building from source requires Rust 1.88 or newer and a C linker. The project lists macOS, Linux, and Windows as supported platforms.

Found this useful?

Share it with someone who would like ai-session-search.

Comparisons