Installation

RepoDNA is one command-line program, repodna, with the web interface built in. A desktop app is also available. Everything runs on your machine. To analyze a repository or look at analyses without installing anything, use the web version, which analyzes in your browser, without Git history. In GitHub Actions, use the GitHub Action.

• Prebuilt binaries
• The desktop app
• Container image
• Web interface image
• npm packages
• WebAssembly
• GitHub Action
• Build from source
• Check the installation
• Uninstall

Requirements. Git on your PATH for history analysis (RepoDNA works without it, but reports no history). Nothing else: the binaries have no runtime dependencies.

Prebuilt binaries#

Download the archive for your platform from the releases page:

The Linux binaries are statically linked and run on any distribution.

Linux and macOS#

tar -xzf repodna-1.3.2-x86_64-unknown-linux-musl.tar.gz
sudo install -m 0755 repodna-1.3.2-x86_64-unknown-linux-musl/repodna /usr/local/bin/repodna
# or, without sudo:
mkdir -p ~/.local/bin && install -m 0755 repodna-1.3.2-x86_64-unknown-linux-musl/repodna ~/.local/bin/

Windows#

Extract the .zip, move repodna.exe to a folder of your choice (for example %LOCALAPPDATA%\Programs\RepoDNA), and add that folder to your PATH.

Verify the download#

Each release has a SHA256SUMS.txt:

sha256sum --check --ignore-missing SHA256SUMS.txt           # Linux
shasum -a 256 --check --ignore-missing SHA256SUMS.txt       # macOS

Get-FileHash .\repodna-1.3.2-x86_64-pc-windows-msvc.zip -Algorithm SHA256   # compare with SHA256SUMS.txt

Unsigned binaries#

The binaries and installers are not signed with a developer certificate.

• macOS blocks programs downloaded from the internet that Apple has not checked. For the command line, remove the quarantine attribute: xattr -d com.apple.quarantine /usr/local/bin/repodna. For the desktop app, open it once and close the warning, then open System Settings > Privacy & Security, choose Open Anyway next to the message about RepoDNA, and confirm. On macOS 14 and earlier, you can instead Control-click the app in Finder and choose Open.
• Windows SmartScreen may say "Windows protected your PC". Choose More info, then Run anyway.

The desktop app#

Installers for Linux (.deb, .rpm), macOS (.dmg), and Windows (.msi, .exe) are attached to each release, and from 1.3.1 on for Windows on Arm (_arm64_en-US.msi, _arm64-setup.exe). See the desktop app.

Container image#

Each release is also published as a container image with the command line and Git, for CI jobs and machines where you would rather not install anything. It runs on linux/amd64 and linux/arm64:

docker run --rm -v "$PWD:/work" ghcr.io/sanskarin/repodna analyze .
docker run --rm -v "$PWD:/work" --user "$(id -u):$(id -g)" \
  ghcr.io/sanskarin/repodna report . --output repodna-report

The current directory is mounted as /work, and --user makes the files RepoDNA writes yours. Tags follow the releases: 1.3.2, 1.3, 1, and latest. Stored analyses live in /tmp/repodna inside the container and disappear with it; mount a volume there (-v repodna-data:/tmp/repodna) to keep them. repodna serve in a container listens on the container's own loopback address, so use an installed binary, the desktop app, or the web interface image for the web interface. The image is based on Alpine Linux; the Alpine packages in it, such as Git, keep their own licenses, and their sources are available from Alpine Linux.

Web interface image#

The web interface is also published as a container image, for teams that want the web version on their own network or without an internet connection. It runs on linux/amd64 and linux/arm64:

docker run --rm -p 8080:8080 ghcr.io/sanskarin/repodna-web

Then open http://localhost:8080. Like the web version, it opens analyses from files in the browser (repodna analyze . --format json > repodna.json, or a .repodna export) and never uploads them; the server only sends the interface. It cannot analyze repositories itself: use the command line or the desktop app for that. nginx serves it as an unprivileged user on port 8080, with the same security headers as repodna serve. Tags follow the releases, as for the command line image.

npm packages#

With Node.js 18 or later, npm installs the command line:

npm install --global @sanskarin/repodna
repodna --version

The packages are on npmjs.com and on GitHub Packages from 1.2.0 on. Install the command line with optional dependencies, which npm includes by default: --omit=optional leaves out the platform package and the command cannot start. The binaries are the same as the prebuilt binaries, so the notes about unsigned binaries apply here too.

From GitHub Packages#

GitHub Packages asks for a token even to install public packages. Create a personal access token (classic) with the read:packages scope, then send the @sanskarin scope there:

npm config set @sanskarin:registry https://npm.pkg.github.com
npm config set //npm.pkg.github.com/:_authToken YOUR_TOKEN
npm install --global @sanskarin/repodna

npm config delete @sanskarin:registry goes back to npmjs.com.

WebAssembly#

From 1.3.1 on, RepoDNA's analysis, reports, and Project DNA cards are also built as WebAssembly, the program the web version runs in the browser. It runs wherever Node.js or a WASI runtime does, with nothing native to install, but without Git it analyzes no history.

# With Node.js
npx @sanskarin/repodna-wasm analyze path/to/repo --output repo.repodna
npx @sanskarin/repodna-wasm report repo.repodna --format html --output report.html

# With a WASI runtime, from the release's repodna-<version>-wasm32-wasip1.zip. The folder
# is mapped to /my-project, whose name the analysis takes.
wasmtime run --dir .::/my-project repodna.wasm analyze /my-project > my-project.repodna

In JavaScript, @sanskarin/repodna-wasm/node analyzes a folder or an archive on disk, and @sanskarin/repodna-wasm runs the program in a browser on files you give it; see the package's README.

Build from source#

Prerequisites:

• Git
• A current stable Rust toolchain (install with rustup)
• Node.js 20.19 or newer with npm, to build the web interface

git clone https://github.com/sanskarIN/RepoDNA.git
cd RepoDNA
npm ci                                   # web interface dependencies
npm run build -w @repodna/web            # builds apps/web/dist, which the binary embeds
cargo install --path crates/repodna-cli --locked

cargo install puts repodna in ~/.cargo/bin, which rustup adds to your PATH.

Without Node.js, skip the two npm commands: everything works except the interactive web interface of repodna serve, which then shows a basic page that lists stored analyses and their reports. You can also install directly from Git this way:

cargo install --git https://github.com/sanskarIN/RepoDNA --tag v1.3.2 --locked repodna-cli

To build the desktop app, see apps/desktop/README.md. For development builds, see development.

Shell completions#

repodna completions bash > ~/.local/share/bash-completion/completions/repodna
repodna completions zsh > "${fpath[1]}/_repodna"
repodna completions fish > ~/.config/fish/completions/repodna.fish
repodna completions powershell >> $PROFILE

Check the installation#

repodna --version
repodna doctor

repodna doctor checks storage, the cache, Git, the language definitions, your configuration, plugins, and AI settings, and says what to do about anything that is wrong.

Uninstall#

Delete the repodna binary (or run cargo uninstall repodna-cli), and remove the data and configuration directories listed in privacy if you want to delete stored analyses and settings too.