Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Emanuele-web04/synara/llms.txt

Use this file to discover all available pages before exploring further.

This guide walks you through the end-to-end flow of getting Synara running and sending your first message to an AI coding agent. By the end you will have Synara open, a local Git repository registered as a project, a provider connected, and an active agent session showing chat output alongside a live terminal.
1

Install Synara

If you have not already installed Synara, download the desktop app for your platform or build from source:Launch the app once installation is complete. Synara starts a local server on localhost:3773 and opens the workspace window.
2

Open or create a project

Synara organizes work around projects, each of which points at a directory on your machine.
  1. Click New Project in the sidebar (or use the project picker at the top of the project list).
  2. Choose a local directory that contains a Git repository. Synara reads the repo’s branch and worktree state from that path.
  3. Give the project a name and confirm.
The project now appears in the sidebar. Synara will use this directory as the working root for any agent sessions and terminals you open under it.
Synara works best with Git repositories. Some features — diffs, branch creation, commit, push, and worktree isolation — require a valid Git repo at the project root.
3

Choose and configure a provider

Synara supports Claude Code, Codex, Gemini, Grok, OpenCode, Cursor, Kilo Code, and Pi. For your first session, Claude Code or Gemini are straightforward choices because they do not require a separate CLI installation.
  1. Open Settings from the sidebar or the gear icon.
  2. Navigate to the Providers section.
  3. Select the provider you want to use and follow the on-screen instructions to connect your account or enter credentials.
Codex requires the Codex CLI to be installed and authorized on your machine before it will work in Synara. See the Installation guide for details.
Once your provider is connected, close Settings and return to your project.
4

Start a chat thread and send a message

With your project open and a provider configured:
  1. Click New Thread inside your project.
  2. Select your provider from the model picker at the bottom of the composer.
  3. Type a prompt — for example, ask the agent to explain a file, write a function, or fix a bug — and press Enter (or click the send button).
Synara forwards your message to the provider and begins streaming the response into the chat timeline.
5

Watch the agent work in split view

As the agent responds, Synara surfaces everything in a unified split layout:
  • Chat timeline — the streamed assistant response, tool calls, and your conversation history.
  • Terminal — a live PTY terminal showing any shell commands the agent runs, file writes, and process output in real time.
  • Diff panel — changed files appear as diffs as the agent edits your codebase. You can review, accept, or navigate changes from this panel.
When the agent’s turn completes, you can reply to continue the conversation, commit the changes directly from the Git controls in the app, or open a branch and hand the thread off to a different provider.
Run agents in parallel with Git worktrees. If you want a second agent working on a different feature at the same time, create a new thread and point it at a separate Git worktree. Worktrees give each session its own isolated working tree on disk so the two agents never touch the same files or conflict on branch state. Look for the worktree picker when starting a new thread.

​
Next steps

Parallel Agents

Run multiple agent sessions across projects and worktrees at the same time.

Provider Handoff

Pass an active thread to a different AI provider mid-session with shared context.

Git Workflow

Review diffs, create branches, commit, push, and open PRs from inside Synara.

Configuration

Customize ports, data directories, keybindings, and provider settings.

Build docs developers (and LLMs) love

Get started for freeTalk to us