real-time agent activity state visualization with character animation
Monitors Claude Code CLI process output in real-time and maps agent execution states (typing, reading files, running commands, waiting for input) to animated pixel art character animations displayed in a persistent office environment. Uses terminal output parsing to infer agent state transitions and triggers corresponding sprite animations without direct API access to the Claude Code process.
Unique: Uses terminal output parsing to infer multi-agent state without direct API integration, rendering state as animated pixel art characters in a persistent office metaphor — a visualization-first approach that treats agent monitoring as a game-like experience rather than a technical dashboard
vs alternatives: Provides visual, gamified agent monitoring that's more engaging than raw terminal logs, while requiring no changes to existing Claude Code workflows or API integration
multi-agent spawning and lifecycle management via ui
Provides a UI button ('+Agent') to spawn new Claude Code CLI terminals with configurable launch options, manages agent lifecycle (creation, termination, reassignment), and persists agent desk assignments across VS Code sessions. Integrates with VS Code's terminal system to create isolated agent processes while maintaining a visual registry of all active agents in the office environment.
Unique: Wraps Claude Code CLI spawning in a game-like office UI where agents are assigned to desks, persisting layout state across sessions — treating agent management as spatial organization rather than a command-line task
vs alternatives: Reduces friction for spawning multiple agents compared to manual CLI invocation, while providing persistent visual organization that survives VS Code restarts
permission bypass configuration via context menu
Exposes a right-click context menu option on agents to launch with the '--dangerously-skip-permissions' flag, bypassing Claude Code's tool approval prompts. This is a direct pass-through to the Claude Code CLI flag system, allowing developers to skip interactive permission dialogs for agents that have been pre-approved or are running in trusted environments.
Unique: Exposes a dangerous-by-design CLI flag through a UI context menu, making permission bypass discoverable but clearly marked as risky — a transparency-first approach to security configuration
vs alternatives: Provides one-click permission bypass for trusted workflows without requiring manual CLI flag entry, though with clear naming that signals the security implications
persistent office layout design and state management
Provides an interactive office editor where developers can customize floor colors (HSB controls), wall colors with auto-tiling, grid-based desk placement (up to 64×64 tiles), and character desk assignments. Layouts are persisted as JSON files and shared across all VS Code windows in a workspace, enabling consistent visual organization of agents across sessions and team collaboration through layout file sharing.
Unique: Treats agent organization as spatial office design with persistent JSON state that survives restarts and can be shared across developers — a metaphor-driven approach to agent registry management that prioritizes visual organization over functional configuration
vs alternatives: Provides a more engaging and team-shareable way to organize agents compared to flat agent lists, though with no functional impact on agent execution
sub-agent visualization for task tool decomposition
Automatically detects when Claude Code agents spawn sub-agents via the Task tool and visualizes these hierarchical relationships in the office environment. Sub-agents appear as additional characters, allowing developers to see the full tree of agent decomposition and understand how complex tasks are being broken down into parallel or sequential sub-tasks.
Unique: Automatically detects and visualizes Task tool sub-agent spawning without explicit configuration, rendering hierarchical agent relationships as a flat office scene where sub-agents appear as additional characters
vs alternatives: Provides automatic visibility into agent decomposition without requiring manual configuration, though with limited insight into task dependencies or execution order
optional audio notifications on agent completion
Provides a toggleable audio notification system that plays a sound when agents complete their tasks or reach terminal states. Notifications can be enabled/disabled via extension settings, allowing developers to receive auditory feedback without constantly monitoring the visual office display.
Unique: Provides simple binary audio notification toggle without granular control or customization — a minimal approach to auditory feedback that prioritizes simplicity over flexibility
vs alternatives: Offers basic audio notifications for agent completion with minimal configuration overhead, though lacking the granularity of more sophisticated notification systems
workspace-persistent agent registry with cross-window synchronization
Maintains a persistent registry of all spawned agents and their desk assignments that survives VS Code restarts and is automatically synchronized across all VS Code windows in the same workspace. Agent state is stored as JSON in workspace settings, enabling consistent agent organization and visibility regardless of which window a developer is working in.
Unique: Stores agent registry and desk assignments in VS Code workspace settings with automatic cross-window synchronization, leveraging VS Code's built-in state persistence rather than external databases
vs alternatives: Provides simple, zero-configuration persistence that works across VS Code windows without requiring external state management, though with limited conflict resolution and no version history
open-source asset system with custom asset directory support
Provides a modular asset system for pixel art characters, furniture, floors, and walls using open-source JIK-A-4 Metro City artwork. Developers can extend the asset library by adding custom assets from local filesystem directories, allowing teams to create branded or themed office environments without modifying the extension code.
Unique: Provides an open-source asset system based on JIK-A-4 Metro City artwork with support for custom local asset directories, enabling community contributions and team customization without requiring extension code changes
vs alternatives: Allows visual customization through asset swapping without modifying extension code, though with undocumented asset format and no built-in asset management tools
+1 more capabilities