pairwise-tui

A keyboard-driven terminal UI for QA engineers and developers who want to generate pairwise and combinatorial test cases with Microsoft PICT. Build PICT models, define constraints, generate high-coverage test suites, and export results on Linux or Windows.

Live site • Releases • PICT

PICT-powered • Linux and Windows binaries • TXT / JSON / CSV / XLSX export

Why use pairwise-tui?

• Generate pairwise and higher-order combinatorial test cases with Microsoft PICT
• Build and edit PICT-compatible models from a fast terminal UI
• Add constraints without leaving the keyboard
• Export test cases to TXT, JSON, CSV, or XLSX
• Use AI assistance to draft parameters and constraints faster

Download & Run

Pre-built binaries for Linux and Windows are available on the Releases page — no runtime required.

1. Download the binary for your platform from the latest release.
2. Run it: ./pairwise-tui.exe on Windows or ./pairwise-tui on Linux.

Windows SmartScreen warning

On first launch Windows may show "Windows protected your PC". This happens because the binary is not yet known to Microsoft's reputation system — it is not a malware detection.

To run the app:

1. Click More info in the SmartScreen dialog.
2. Click Run anyway.

Alternatively: right-click the .exe → Properties → check Unblock → OK, then run normally.

Interface

Three tabs, navigated with 1 / 2 / 3 or [ / ]:

Tab	Purpose
Model	Define parameters (name + comma-separated values) and PICT constraints
Options	Set combination order, randomize seed, case sensitivity, output file path, output format, storage path, file template, AI model
Results	View generated test cases in a scrollable table

Keyboard shortcuts

Global:

Key	Action
g	Generate test cases (runs PICT)
s	Save results to output file
w	Write model to storage
o	Open model from storage (file picker)
m	Open message log overlay
t	Cycle through themes
?	Open PICT documentation
i	Open AI prompt overlay (or AI setup if no API key configured)
F2	Open AI setup overlay
q / Ctrl+C	Quit
1 / 2 / 3	Switch tabs
[ / ]	Previous / next tab

Model tab — params panel:

Key	Action
a	Add parameter
d	Delete selected parameter
e	Edit values for selected parameter
c	Edit constraints
x	Clear entire model (prompts for confirmation)
Tab	Switch to values panel
↑ / ↓	Navigate parameters
Escape	Return to params panel / cancel edit

Results tab:

Key	Action
↑ / ↓	Scroll test cases

Message log overlay (m):

Key	Action
↑ / ↓	Navigate log entries
c	Copy selected entry to clipboard
a	Copy all entries to clipboard
m / Escape	Close overlay

AI setup overlay (F2 or i without key):

Key	Action
Enter	Save API key
d	Clear saved API key
Escape	Close overlay

AI prompt overlay (i):

Key	Action
Ctrl+G	Generate parameters from description
F2	Switch to AI setup
Escape	Cancel

Documentation overlay (?):

Key	Action
↑ / ↓	Navigate chapters / scroll content
Enter	Open selected chapter
Escape	Back / close

AI Parameter Generation

Pairwise-TUI can generate PICT parameters automatically from a natural language description using Claude.

Setup

1. Get an Anthropic API key from console.anthropic.com.
2. Press F2 (or press i when no key is configured) to open the AI Setup overlay.
3. Type your API key and press Enter to save it. Press d to clear a saved key.

The key is stored at:

• Linux/macOS: ~/.config/pairwise-tui/credentials.json (chmod 600)
• Windows: %APPDATA%\pairwise-tui\credentials.json

The ANTHROPIC_API_KEY environment variable takes precedence over the stored key.

Usage

1. Press i to open the AI prompt overlay.
2. Describe the feature or test scenario in plain text, e.g. Login form with email, password, and remember-me checkbox.
3. Press Ctrl+G to generate parameters.
4. Review the generated parameters in the Model tab.
5. Press g to run PICT and generate test cases.

Model selection

In the Options tab, tab to the AI Model field and press Enter to cycle through models:

Model	Characteristics
claude-haiku-4-5	Fastest, lowest cost (default)
claude-sonnet-4-6	Balanced speed and capability
claude-opus-4-6	Most capable, highest cost

Model file format

The model is saved/loaded as a plain PICT text file (model.txt):

OS: Windows, Linux, macOS
Browser: Chrome, Firefox, Safari
Language: EN, DE, FR

IF [OS] = "Linux" THEN [Browser] <> "Safari";

Development

Requirements

• Bun runtime
• Git LFS (for PICT binaries)

The PICT binaries (binaries/pict and binaries/pict.exe) are stored in Git LFS. You must have Git LFS installed and pull the files before running the app:

git lfs install
git lfs checkout

Then install dependencies and start:

bun install
bun run dev

Available scripts

Command	Description
bun run dev	Run in watch mode
bun run start	Run the app
bun run check	Biome lint/format check
bun run typecheck	TypeScript type-check
bun run build:dist	Build Linux and Windows executables to dist/
bun run build:dist:linux	Build the Linux executable
bun run build:dist:windows	Build the Windows executable
bun run package:appimage	Package the Linux AppImage

Building standalone executables

bun run build:dist

Outputs self-contained executables to dist/ for Linux and Windows (no Bun required to run).

Windows builds must run on Windows for the custom executable icon to be embedded.

To package the Linux AppImage after building the Linux executable:

bun run build:dist:linux
APPIMAGETOOL=./dist/appimagetool.AppImage bun run package:appimage

The release workflow publishes three artifacts:

• pairwise-tui.exe for Windows with the custom app icon
• pairwise-tui as the raw Linux executable
• Pairwise-TUI-x86_64.AppImage as the desktop-friendly Linux download

To create a new release:

git tag v1.2.3
git push --tags

This triggers the GitHub Actions release workflow, which builds the Windows executable, Linux executable, and Linux AppImage, then publishes them to the Releases page automatically.