All projects

Planned

Ferrost / Ferrostide

Rust-native workflow experiments intended to reduce context switching between the editor, terminal, and AI tools.

Project overview

Overview

Ferrost and Ferrostide are workflow-consolidation experiments aimed at reducing the context switching between an editor, terminal, and browser-based AI tools.

The near-term goal is not to replace all of those applications. It is to build a useful Rust-native project workbench that can grow in layers.

Project structure

Name roles

  • Ferrost is the terminal and workbench foundation.
  • Ferrostide is the IDE and workspace mode built on top of Ferrost.

Rationale

Why this experiment matters

The project targets a workflow used every day, making it closer to immediate practical value than deeper desktop or compositor work.

It also creates a natural integration point for Ferrule, Ferroforge, Feo, and Ferrisle.

Foundation

Ferrost MVP

Ferrost opens a project workspace, understands basic repository status, runs project commands, and reduces at least one meaningful context switch.

  • Open a project path.
  • Show the current project name.
  • Show the Git branch and status.
  • Run configured commands.
  • Save useful project commands.
  • Integrate with Ferrule later.
  • Launch external editors, terminals, or browsers at first when that keeps the scope smaller.

Workspace mode

Ferrostide MVP

Ferrostide adds focused workspace and IDE capabilities on top of the Ferrost foundation.

  • A project tree.
  • Basic editing.
  • A terminal or session area.
  • An AI sidecar placeholder.

Long-term direction

Possible future features

  • File tree.
  • Text editor pane.
  • Terminal pane.
  • Git and status pane.
  • AI or chat sidecar.
  • Project memory and context.
  • Ferrule integration.
  • Ferroforge integration.
  • Feo and Ferrisle-aware helpers.

Scope control

Each layer must be useful by itself

The project should not begin as a terminal emulator, editor, browser, AI client, Git client, file manager, and desktop shell all at once.

Each layer needs to solve a real workflow problem before the next layer is added.

Success criteria

First useful version

  • Open a project.
  • Show project status.
  • Run project commands.
  • Reduce at least one real context switch.
  • Explain the long-term vision and current limited scope in the README.

Boundaries

Non-goals for now

  • A terminal emulator built from scratch.
  • A complete VS Code replacement.
  • A full browser.
  • A full ChatGPT replacement.
  • Complex UI framework debates before the workflow is clear.
  • Desktop-environment scope creep.

Next action

Next step

Write a short VISION.md or README.md that defines the smallest daily workflow Ferrost should improve first.