Skip to content

1.0.0 - First release!

Choose a tag to compare

@kfiramar kfiramar released this 11 Mar 22:25
· 1584 commits to main since this release
515819a

envctl 1.0.0

envctl is a CLI for spinning up full local development environments across a main repo and multiple worktrees without manual port juggling, dependency collisions, or repetitive setup work.

This first release focuses on one core idea: local multi-environment development should feel deterministic. envctl lets you bring up backend, frontend, and supporting dependencies in parallel, isolate each environment cleanly, and operate all of them through one command surface.

Highlights

  • Start full local environments for a main repo and many worktrees
  • Automatic port allocation to avoid backend/frontend/dependency conflicts
  • Isolated dependency orchestration for services like Postgres, Redis, Supabase, and n8n
  • Unified interactive dashboard for status, health, logs, errors, restart, and actions
  • Built-in workflows for testing, review, commit, PR creation, and migrations
  • First-run configuration wizard with managed .envctl setup
  • Resume and restore support for existing runs
  • Per-run logs, artifacts, failure summaries, and debug bundles
  • AI prompt preset installation for Codex, Claude, and OpenCode
  • Strong support for high-throughput worktree-based development and AI-assisted implementation

Why It Exists

Modern local development gets messy fast when you work on multiple branches or worktrees at once. Ports collide. Shared services stomp on each other. App env vars drift. Logs scatter across terminals. Test and review flows become inconsistent.

envctl solves that by acting as an orchestration layer for your local repo:

  • it decides what should run
  • allocates ports consistently
  • wires the right env for each environment
  • keeps dependencies isolated
  • gives you one place to inspect and operate everything

Included In 1.0.0

  • Python runtime engine
  • Interactive and headless command flows
  • Worktree-aware planning and execution
  • Textual-based UI flows for selectors and config
  • Prompt installation support for built-in AI workflows
  • Packaged wheel and source distribution for installation