mr-manager 1.0.0

TUI-based CLI for managing repositories inside a *myrepos* config file. Simplifies adding and removing repos with an interactive interface.

mr-manager

---

mr-manager is a Textual TUI for managing repositories in your ~/.mrconfig myrepos file. It discovers Git repositories under your home directory and lets you toggle which ones are tracked.

Features

• Discovers local Git repositories recursively.
• Fast startup using a cached filesystem scan (configurable TTL).
• Reads existing repo sections from ~/.mrconfig.
• Interactive keyboard-driven selection UI.
• Saves add/remove config updates safely.
• Highlights missing configured repositories.
• User-configurable discovery root and cache TTL persisted in ~/.config/mr-manager/config.yaml.

Installation

Python (pipx)

Requires Python >=3.13.

pipx install mr-manager

From source

Prerequisites: Python >=3.13, uv

uv sync --dev
uv run mr-manager

Usage

• mr-manager: Start the interactive TUI
• mr-manager -v: Print the installed version and exit

Keybindings

• space: Toggle selected repository
• j: Move down
• k: Move up
• s: Save changes
• r: Refresh repository scan (bypasses cache)
• c: Open configuration editor
• q/ESC: Quit (asks for confirmation if unsaved changes exist)

User configuration

• Config file path: ~/.config/mr-manager/config.yaml
• When missing, defaults are used and the file is created when you save from the config editor.
• Currently configurable:
  • Discovery Cache TTL (hours)
  • Discovery Root Directory

Development workflow

Run checks locally before opening a PR:

uv run ruff check --no-fix .
uv run ruff format --check .
uv run ty check
markdownlint --config markdownlint.json --ignore-path .markdownlintignore "**/*.md"

Performance Benchmarking

The project includes a benchmarking script to track performance regressions across versions:

# Run a quick benchmark
uv run scripts/benchmark.py --runs 5

The benchmark supports step-based execution and persistent artifacts for later analysis:

# Full local run (collect + text summary + graph)
uv run scripts/benchmark.py --runs 5

# CI-style text-only run
uv run scripts/benchmark.py --runs 10 --steps collect,summary

Default artifacts are saved in .cache/mr-manager/benchmarks/<benchmark-id>/:

• data.json raw measurements
• summary.txt textual report
• plot.png graph (when plot step is enabled)

Useful options:

• --versions v0.0.1 main to benchmark specific refs
• --scan-root <path> to control discovery root
• --steps collect,summary,plot to resume from previous data
• --force-collect to overwrite existing data.json

Documentation

Project docs are maintained in the wiki/ folder and published to GitHub Wiki via CI.

• Wiki Home
• Contribution Guidelines
• Troubleshooting
• Release Process