Alita

Developers juggle a dozen tools to answer simple questions: where did today go, which app is eating the bandwidth, why is the machine slow, which pull requests are waiting for me. Most trackers send that data to someone else’s cloud, which many companies cannot allow, and most system tools change things without asking.

1. 01

   Local first, observe only

   Data lives on the machine in SQLite shards and secrets are encrypted with the operating system’s keystore. Nothing is sent anywhere except the addresses you configure, and killing a process, deleting files, changing a firewall rule or writing to Jira always asks first.

2. 02

   One file, small parts

   The app is bundled into a single file from small source parts in strict layers, from core and platform to tracking, monitors, engine, server and UI, and tests enforce the layering. Installers for all three platforms support silent installs and organisation defaults for rollout.

3. 03

   Native where it matters

   A small background agent in Rust and Go keeps measuring when the window is closed, a Go front door checks every request before the Python hub sees it, and a Rust core, also compiled to WebAssembly, powers the formatters, validators, diff and OpenAPI tools.

4. 04

   A complete API workspace

   Collections, environments, scripts and tests, a collection runner, Postman import and export, OpenAPI tools with a local mock server, and a Rust HTTP engine that handles proxies, company certificate authorities and client certificates.

5. 05

   Delivery and safety in view

   A Git board with pull and merge request pipelines, secret scanning, dependency vulnerabilities from OSV, runtime end of life dates and certificate expiry, with a separate notification choice for every topic.

Local first changed every decision. Background jobs keep their last result so pages open instantly without the internet, and nothing that changes the system happens without a yes.

Python, TypeScript, Rust, Go, SQLite, WebAssembly