Overview
Understand how Agentty organizes AI agent sessions and workflows.
agentty is an ADE (Agentic Development Environment) for AI-assisted software development in your terminal.
It runs AI coding agents in dedicated AI sessions.
What Agentty Provides🔗
When you start a session, Agentty:
- Spawns a clean worktree branch for every live session.
- Runs agent-driven edits in isolation from your base branch.
- Keeps terminal output, diffs, and generated changes in one reviewable stream.
- Keeps already-published session branches synced after later completed turns.
Typical Flow🔗
A session takes your prompt through agent-driven edits to a reviewable diff. You decide whether to merge or discard the changes.
Follow Installation for your first session. Sessions can also be staged as drafts or stacked on top of another session; see Workflow for those options.
Worktree Isolation🔗
Every session runs in its own git worktree, created automatically when the live session starts:
The session starts from your local active branch, even if it is behind the remote. Sync the project first to include remote changes. Agentty checks the worktree before each turn and keeps session edits separate from your base branch until you merge. Bare-repository projects are also supported.
Worktrees are stored under ~/.agentty/wt/ and are cleaned up automatically when a session reaches Done or Canceled, or when you delete a session.
Auto-Update🔗
Agentty checks for updates at startup and hourly. Installed agent CLIs are also refreshed at startup. See Auto-Update for status messages and controls.
Key Concepts🔗
- Agent: An external AI CLI backend (Antigravity, Claude, Codex, or Gemini) that performs coding work. See Agents & Models.
- Session: An isolated unit of work: a conversation, a worktree branch, and a reviewable diff. See Workflow and Keybindings.
- Project: A git repository registered in Agentty. Select between projects with the Projects tab.
- Diff view: The changes an agent made in a session, ready for your review.
Development Status🔗
Agentty is in active development. Releases follow Semantic Versioning, but the current 0.y.z series may still introduce breaking changes as workflows, integrations, and safeguards evolve. Always review and verify changes before relying on them in your repositories.
Next Steps🔗
- Installation — install Agentty and run it for the first time.
- Agents & Models — configure backends and choose models.
- Workflow — learn the interface layout and session lifecycle.
- Keybindings — learn the keyboard shortcuts for each view.