Local-first search over your Claude and Codex session history, with a Rust search core, JSON CLI, and desktop app.
yjid/kensaku
A free, local, open-source desktop launcher that searches file names, text inside documents, OCR'd screenshots, and meaning, with MCP access for AI agents.
Open-source alternative to:Everything
Listary
- Stars
- 1,023
- Forks
- 72
- Open issues
- 13
- Last commit
- yesterday
What is it?
Kensaku is a desktop file launcher for Windows, macOS, and Linux. It indexes the folders you choose and searches by file name, by text inside PDFs, DOCX, EPUB, code and logs, by text read from images via OCR, and by meaning. Results update as you type, and you can use natural-language time filters such as "last week" or a query syntax with operators like type:pdf and modified:7d.
The app runs entirely on your machine with no account, telemetry, or cloud search. Filename search over about 100,000 files is reported at around 1.3 ms in memory, with an index of roughly 12 MB for names and metadata. It starts hidden in the system tray and is summoned with a keystroke. A command-line tool, kensaku-cli, ships with each release.
Kensaku also exposes the local index to AI agents over MCP, read-only by default, and offers a loopback HTTP server protected by a bearer token.
Key features
- 01Searches file names, full text inside documents, OCR text in images, and meaning in one query.
- 02Reads PDF, DOCX, PPTX, XLSX, EPUB, RTF, code, logs, and CJK text.
- 03Supports natural-language time filters and a query syntax with operators.
- 04Indexes incrementally, picking up changes without rescanning the whole disk.
- 05Runs as a tray launcher with a global hotkey and a Quick Look-style preview.
- 06Exposes the index to AI agents over MCP in read-only mode.
Use cases
- Finding a PDF or note when you only remember its contents, not its filename.
- Searching error messages inside screenshots or log files.
- Giving a local AI agent read-only search access to your files without sending data to the cloud.
Quick start
- 1
Download the bundle for your system from the latest release page
- 2
Install the CLI with a search example
kensaku-cli search "docker error" - 3
Build from source: clone the repository
git clone https://github.com/yjid/kensaku - 4
Install dependencies and build the desktop app
bun install && bun run tauri build
Things to know
- Builds are not yet code-signed, so macOS and Windows show a warning on first run.
- Building from source requires Rust, Bun 1.1 or later, and platform webview libraries.
- The auto-updater is not wired up yet.
- Windows has received the most real-world testing so far; macOS and Linux have less.
In collections
Topics
FAQ
- What is kensaku?
- A free, local, open-source desktop launcher that searches file names, text inside documents, OCR'd screenshots, and meaning, with MCP access for AI agents. Kensaku is a desktop file launcher for Windows, macOS, and Linux. It indexes the folders you choose and searches by file name, by text inside PDFs, DOCX, EPUB, code and logs, by text read from images via OCR, and by meaning. Results update as you type, and you can use natural-language time filters such as "last week" or a query syntax with operators like type:pdf and modified:7d.
- Is kensaku free to use? What license does it have?
- kensaku is open source under the MIT license. Check the license terms before using it in commercial products.
- Is kensaku still maintained?
- Actively maintained. The last commit to kensaku was on Nov 6, 2026.
- What can I use kensaku for?
- Finding a PDF or note when you only remember its contents, not its filename. Searching error messages inside screenshots or log files. Giving a local AI agent read-only search access to your files without sending data to the cloud.
- How do I install kensaku?
- Download the bundle for your system from the latest release page Install the CLI with a search example: kensaku-cli search "docker error" Build from source: clone the repository: git clone https://github.com/yjid/kensaku Install dependencies and build the desktop app: bun install && bun run tauri build
- Where can I get kensaku?
- The source code and installation instructions are on GitHub: https://github.com/yjid/kensaku
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Local-first search CLI and agent tool that combines ripgrep, BM25 and vector search over your workspace.
A private, native macOS app that captures pages, text, links, notes and images, and finds them again with full-text search and OCR.
Open-source desktop agent app and Rust library for running many AI agents with lower token use, memory and resource cost.
Desktop app for screenshots, screen recording, and local OCR on macOS, Windows, and Linux, with assets stored on your machine.
A local-first notebook environment where humans, Jupyter kernels, and AI agents edit the same live document, with a native desktop app.