Getting Started

Most developers using AI coding assistants run them in loose terminal sessions. Context is lost between runs, there is no history of what was tried, and coordinating multiple agents across branches is manual. AgentsInFlow solves this by sitting on top of your existing AI CLI subscriptions and providing:

You do not need to change your AI workflow. AgentsInFlow wraps whatever CLI you already use and adds the structure around it. The screenshot below shows the assistant chat view with conversation history.

• The Assistant view provides a full-screen chat interface for interacting with AI agents.
• Conversations are persisted across sessions and can be organized into folders in the sidebar.
• The engine and model selector at the bottom lets you switch between Claude, Codex, and Cursor mid-conversation.

System Requirements

Download and Install

Authenticate Your AI CLI

AgentsInFlow does not manage AI credentials directly — it delegates to the CLI you have installed. Before running your first ticket, make sure at least one CLI is authenticated:

npm install -g @anthropic-ai/claude-code

claude login

npm install -g @openai/codex

codex login

cursor-agent

cursor-agent --login

AgentsInFlow’s interface is organized into three main regions that work together to give you visibility into every layer of your project.

Sidebar

The Sidebar occupies the left edge of the window (280px default width). It contains:

In collapsed mode, the sidebar shows project icons with color-coded badges. Each icon represents a project you can switch to with a single click.

Content Area

The Content Area fills the center of the window and displays whatever view mode you are in — workspace tabs, kanban board, dashboard, git history, activity log, or the full assistant chat.

Inspector Panel

When you open a ticket in Workspace mode, the content area splits into two resizable panels:

The left panel shows the ticket description and metadata. Below is the ticket details panel for a running execution.

• The ticket title and type badge appear at the top.
• The markdown description fills the body of the panel.
• Tabs for Children and Depends On let you manage ticket relationships.

The right panel displays the execution terminal. Below is a running agent session with the config bar and live output.

• The config bar at the top shows the engine, model, branching options, and live token metrics.
• The terminal below streams agent actions in real time.
• Buttons for Stop, Commit, Merge, and Mark as Done appear in the bar.

Drag the divider between panels to resize. The split persists across sessions.

AgentsInFlow has six view modes, each accessible via keyboard shortcut or the ViewModeSwitcher buttons at the top of the sidebar.

The view mode buttons are always visible at the top of the window. The active mode is highlighted with the project’s accent color.

• Each button shows an icon and label for the mode.
• The active mode (here Workspace) is highlighted with the project accent color.
• Click any button or use the keyboard shortcut (Cmd+1 through Cmd+6) to switch.

The Kanban view organizes tickets into columns by status, giving you a board-level view of work across your project.

• Tickets are grouped into Backlog, Ready, Running, Blocked, Approval, and Done columns.
• Each ticket card shows its type badge, title, and creation date.
• The Active Agents panel on the right shows currently running executions across projects.

The Activity view provides a unified event log of all agent and system actions across your project.

• Events are listed chronologically with timestamps, source type (user, agent, system), and descriptions.
• Filter by event source to isolate agent actions from user actions.

AgentsInFlow organizes work into three layers. Understanding this mental model helps you use the right tool at the right time.

How Work Flows Between Layers

The typical progression is: conversation → ticket → execution → files.

The Dashboard view lets you create custom status boards to visualize your project. It sits alongside the three work layers as a monitoring surface.

• Dashboards use a 12-column grid where you place chart widgets (ECharts), diagram widgets, and markdown panels.
• The Active Agents panel on the right shows all currently running executions with their status and duration.

Not every conversation needs to become a ticket, and not every ticket needs agent execution. Use the layer that matches the formality of your work.

Follow these steps to go from a fresh install to a completed AI-driven task.

5Watch It Work

The right panel shows live terminal output as the agent works. The execution config bar updates in real time with metrics:

Input tokens

emerald badge

Output tokens

sky badge

Context usage

color-coded

Duration

live timer

Context usage is color-coded: green under 50%, amber 50–80%, red above 80%. You can Stop the execution at any time, or let it run to completion.

The execution config bar displays live metrics while the agent runs. Below is the config bar from an active Codex execution.

• The top row shows the attempt timestamp. The middle row has the Engine (Codex), Model (5.4 Extra High), Options, and branch config (from: main, use branch toggle).
• The bottom row shows live metrics: IN (input tokens), OUT (output tokens), CTX (context usage at 9.0%), TIMER, and session ID.

These shortcuts work globally across all view modes.

For the complete shortcut reference and customization options, see Keyboard Shortcuts.

Now that you have the fundamentals, explore the features that matter most to your workflow: