Desktop App for Linux, macOS & Windows

Your AI Coding
Workspace

A desktop app where AI coding agents run in real terminals. Generate code, review pull requests, and merge — without switching tools. Works with Claude Code, OpenAI Codex, and OpenCode, each in its own isolated git worktree.

Start 14-day free trial
14-day free trial No credit card required

Simple, transparent
pricing

14-day free trial
Individual
$20 /mo
or $200/yr (save 2 months)
  • Unlimited agent sessions
  • Git worktree isolation
  • Multi-agent support
  • Issue tracking & PR review
  • See all features
Start 14-day free trial
Team
$30 /seat/mo
or $300/yr (save 2 months)
  • Everything in Individual
  • Per-seat billing
  • Usage analytics
  • Project config sync
  • Centralized team billing
Get started

Isolated Sessions

Every session gets its own git worktree and branch. Multiple agents can work on the same project simultaneously without interfering with each other or your working directory.

Multi-Agent Support

Run Claude Code, OpenAI Codex, and OpenCode in separate sessions. Each agent runs in its own terminal. Launch guest agents to run multiple agents in the same session.

Full Git Workflow

Review diffs, push branches, create pull requests on GitHub, GitLab, and Bitbucket, resolve merge conflicts, and monitor CI/CD checks. Agents can pull, push, and check pipeline status on their own.

Keyboard-Driven

Vim-style navigation across every view with numeric prefixes, visual mode, and search. Jump Mode gives you instant access to tabs, git actions, sessions, and issues.

Agent Sandbox Pro

Lock agents down to just the project worktree. Sandboxed sessions can't read your credentials, access other projects, or reach unexpected network endpoints.

Excavator

Launch multiple AI agents to scan your project for bugs in the background. Review findings with severity levels, then promote confirmed bugs to your issue tracker.

Your subscriptions, your terms. Taskeract launches agent CLIs directly in real terminals — same binary, same credentials, same TUI you'd get in your own terminal. No API wrappers, no proxies. Your existing Pro and Max subscriptions work exactly as intended.
Our commitment: Taskeract won't disappear. If we're ever unable to continue supporting Taskeract, we'll release it as open source. Your workflow is safe. Learn more →

Three steps to
AI coding in isolation.

1

Configure Agents

Taskeract detects AI coding agents on your system and configures them automatically during the setup wizard.

Claude Code
OpenAI Codex
OpenCode
2

Create Isolated Environment

One click to spin up an isolated git worktree with its own local domain, keeping browser state separate.

worktree: ready
URL: https://warm-fox.myapp.tact
ssl: auto-provisioned
3

Run Agent Sessions

Start coding sessions in isolated worktrees. Review diffs with syntax highlighting, push branches, create PRs, and merge when ready. Sessions persist across restarts.

agent: Implementing auth flow...
status: 14 files changed
preview: https://warm-fox.myapp.tact

Isolated local domains
for every environment

Every environment gets its own local domain, so browser cache, cookies, and storage stay completely isolated between projects and sessions. No more projects on localhost clobbering each other. Each domain is served over HTTPS with auto-provisioned certificates, which is also required by some third-party integrations.

  • Isolated browser cache, cookies & storage
  • Unique domain per project
  • Unique subdomain per environment
  • Auto-provisioned HTTPS certificates
Environments — 3 active
auth-refactor
Claude Code • worktree
● running
bold-elk.myapp.tact
api-tests
OpenAI Codex • worktree
● running
dry-oak.shopapp.tact
ui-redesign
OpenCode • worktree
● stopped
red-ash.dashboard.tact

Start Your 14-Day
Free Trial

Full access to all Individual plan features. Cancel anytime.

Start free trial