Ayeneh

A terminal tool that installs packages from mirrors, benchmarks, and reports results, all with both simple TUI and CLI.

Supported Package Managers

Registry	Package Manager
PyPi	pip, uv
npm	npm

Features

• Install packages from mirrors. Switch to a different mirror if fails.
• Benchmarks mirrors and ranks them by dowload speed.
• Interactive keyboard-only TUI built with ratatui.

Installation

Pickup desired binary from the releases Or build from source:

• Make sure you have Rust (stable) installed via rustup

git clone https://github.com/pasta-engineers/ayeneh.git
cd ayeneh
make
# or
cargo build

Usage

The project ships two binaries:

• ayeneh-cli — non-interactive CLI, suitable for CI pipelines, Docker images, and shell scripting.
• ayeneh-tui — interactive Ratatui application.

Get help:

ayeneh-cli --help
ayeneh-cli pip --help
ayeneh-cli npm --help

Environment Variables

Runtime behavior can be tuned without rebuilding:

Variable	Meaning	Default
AYN_DATA_DIR	Directory holding pypi.json/npm.json	./data
AYN_REPORTS_DIR	Directory for generated reports	auto-detected reports/
AYN_TIMEOUT	Per-attempt HTTP timeout (seconds)	15
AYN_ATTEMPTS	Benchmark attempts per mirror	3
AYN_SCHEDULE_INTERVAL_SECS	Delay between scheduler cycles (seconds)	3600

The numeric variables fall back to their default if unset or given a non-numeric, zero, or negative value.

Launch the TUI:

ayeneh-tui

In TUI, you can:

• Run benchmarks
• Reorder mirrors based on benchmark results, so faster mirrors are listed first in installation processes
• Add/remove mirrors manually CLI features are available in TUI as well.

Install python packages from mirrors:

ayeneh-cli pip install <package>

or install from requirements file:

ayeneh-cli pip install -r requirements.txt

Install npm packages from mirrors:

ayeneh-cli npm install <package>

Run a one-off benchmark from the command line:

ayeneh-cli run pypi
ayeneh-cli run npm

Generate a JSON report for both package managers:

ayeneh-cli report

Run the hourly scheduler (keeps benchmarking forever):

ayeneh-cli schedule

Example Report

reports/2026-07-12_14-30.json:

{
  "package_manager": "pypi",
  "generated_at": "2026-07-12T14:30:00+00:00",
  "results": [
    { "mirror": "https://pypi.org/simple/", "latency": 120, "success": 100.0 },
    { "mirror": "https://mirror.example1/simple/", "latency": 180, "success": 100.0 }
  ],
  "best": "https://pypi.org/simple/"
}

Mirror Lists

pypi.json and npm.json each define the sample package to download during benchmarking and the list of mirrors to test it against. By default, the program reads them from ./data:

{
    "package": "requests",
    "mirrors": [
        "https://pypi.org/simple/",
        "https://pypi.devneeds.ir/simple/",
        "https://package-mirror.liara.ir/repository/pypi/"
    ]
}

Set AYN_DATA_DIR when the registry files are stored elsewhere:

AYN_DATA_DIR=/opt/mirror/registries ayeneh-cli run pypi

Edit these JSON files to add or remove mirrors, or to change the package used for benchmarking. Mirror lists are configured statically — no network scraping is performed to discover them.

Docker

A Python-based image is provided that ships the prebuilt ayeneh-cli binary alongside the mirror registry config, so it can be used directly as a base for installing packages or benchmarking. Build it from the repository root (a release binary must already exist):

cargo build --release -p ayeneh-cli
docker build -f docker/Dockerfile.python -t ayeneh-cli .

Run it, or extend it in your own Dockerfile to install packages from mirrors:

docker run --rm ayeneh-cli ayeneh-cli pip install requests

FROM ayeneh-cli
RUN ayeneh-cli pip install -r requirements.txt

A Node.js-based image is provided for npm. Build it the same way (a release binary must already exist):

docker build -f docker/Dockerfile.npm -t ayeneh-cli-npm .

Run it, or extend it in your own Dockerfile to install packages from npm mirrors:

docker run --rm ayeneh-cli-npm ayeneh-cli npm install lodash

FROM ayeneh-cli-npm
RUN ayeneh-cli npm install lodash express