Terminal UI for Agentic AI!
Project description
agent-terminal-ui
Terminal user interface for AI agents built on agent-utilities.
A Textual-based terminal interface for interacting with AI agents. Connects to an agent-utilities backend via dual protocol support: AG-UI (SSE streaming) and ACP (JSON-RPC + SSE).
[!NOTE] This library is in early development and subject to change.
Features
Core Functionality
- Dual protocol support -- AG-UI (SSE streaming, default) and ACP (JSON-RPC + SSE, opt-in)
- Dynamic workflow sidebar -- discovers graph nodes from sideband events at runtime; nodes are never hardcoded
- Phase labels -- Planning, Discovery, Execution, Validation
- Completed node markers -- checkmarks on finished specialists
- Rich tool execution display -- extensible formatter system for rendering tool calls and results
- Human-in-the-loop tool approval -- modal for confirming sensitive tool calls before execution
- Multi-modal image attachment -- attach images to messages for visual reasoning
- Session management and chat history -- browse and resume previous conversations
- MCP server browser -- inspect connected MCP servers and their tools
User Experience
- Message queuing -- queue messages while agent is processing; related queries are intelligently combined using regex patterns for conjunctions, sequential actions, and similar structure
- Exit confirmation -- modal dialog prevents accidental termination via Ctrl+C or
/exit - Terminal transparency -- UI respects your terminal's transparency settings for seamless integration
- Theme system -- multiple built-in themes (modern_dark, modern_light, nord, gruvbox) with proper color semantics
Commands
- Slash commands -- comprehensive command set for common operations:
/help-- show available commands/clear-- clear the current event log/mcp-- browse connected MCP servers and their tools/history-- browse and select from historical chat sessions/image-- attach images to messages/init-- initialize a new project or workspace/review-- review code and suggest improvements/test-- run tests on the current codebase/search-- search through code and documentation/stats-- show statistics about the current session/cost-- show token and cost tracking information/queue-- show current message queue status/queue:clear-- clear all queued messages/queue:toggle-- enable/disable message queuing/model-- switch between available AI models/theme-- switch between available themes/compact-- compact conversation context to save tokens/diff-- show interactive diff viewer for recent changes/recap-- summarize the session context/fast-- toggle fast mode (Haiku/Flash models)/memory-- manage project memory (AGENTS.md)/agents-- list available specialized agents/add-dir-- add a directory to the agent's working context/exit,/quit-- exit the application with confirmation
Input Prefixes
!-- Direct Bash execution (e.g.,!ls -la)@-- Fuzzy file mention autocomplete (e.g.,@app.py)
Backend Integration
- Unified specialist visibility -- MCP agents and A2A peers appear identically in the workflow sidebar; both emit the same sideband events via the backend's
discover_all_specialists()unified roster - Tool-count telemetry --
tools-boundsideband events includetoolset_count,dev_tools, andmcp_toolsbreakdowns for per-specialist visibility - Real-time token and cost tracking -- integrated in the status line for session monitoring
- Memory Auto-loading -- backend automatically includes
AGENTS.mdandMEMORY.mdin the system prompt for project-aware reasoning.
Usage
Start the agent-utilities backend server, then launch the TUI:
agent-tui
Or run with uv if installed locally:
uv run agent-tui
Keyboard Shortcuts
- Ctrl+C -- Interrupt generation or cancel current operation
- Ctrl+D -- Exit session (with confirmation)
- Ctrl+L -- Clear the event log
- Ctrl+O -- Toggle workflow sidebar
- Ctrl+T -- Toggle task list view
- Ctrl+U -- Clear input buffer
- Ctrl+Y -- Restore cleared input buffer
- Alt+P -- Switch AI model
- Alt+T -- Toggle Extended Thinking (for reasoning models)
- Alt+O -- Toggle Fast Mode
- Ctrl+R -- Reverse history search
- Ctrl+H -- Show help overlay
- Tab -- Navigate between focusable elements
- Esc Esc -- Rewind/Undo (experimental)
Message Queuing
When the agent is processing, your input is automatically queued. The system intelligently combines related queries using patterns like:
- Conjunctions: "and", "also", "plus", "then", "after that"
- Sequential actions: semicolon-separated commands
- Similar structure: same action verbs (fix, add, remove, update, create, delete, implement, refactor)
Example: If you type "fix the bug in app.py" followed by "and add a test for it", these will be combined into a single query.
Environment Variables
| Variable | Default | Purpose |
|---|---|---|
AGENT_URL |
http://localhost:8000 |
Agent server URL |
ENABLE_ACP |
false |
Enable ACP protocol instead of AG-UI |
ACP_URL |
http://localhost:8001 |
ACP server URL (when ACP is enabled) |
Themes
The TUI supports multiple themes that respect terminal transparency:
- modern_dark (default) -- Dark theme with blue accents
- modern_light -- Light theme with proper contrast
- nord -- Nord color palette with frosty aesthetics
- gruvbox -- Gruvbox retro color scheme
Switch themes using /theme <name> command.
Development
Prerequisites
Setup
Create uv virtual environment and install dependencies:
uv sync --frozen --all-groups
To update dependencies (updates the lock file):
uv sync --all-groups
Run formatting, linting, and type checking:
uv run ruff format && uv run ruff check --fix && uv run ty check
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file agent_terminal_ui-0.1.36.tar.gz.
File metadata
- Download URL: agent_terminal_ui-0.1.36.tar.gz
- Upload date:
- Size: 103.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
ff15288bdd98fe0a02717681681b1e9e6a9f689d54485b2d0725b016519304e6
|
|
| MD5 |
3740dbee7134291352ba337758d73f35
|
|
| BLAKE2b-256 |
3e34819343578e4a5913916e970134a1b483a7e24a34dc1cf723dc234c2e1744
|
File details
Details for the file agent_terminal_ui-0.1.36-py3-none-any.whl.
File metadata
- Download URL: agent_terminal_ui-0.1.36-py3-none-any.whl
- Upload date:
- Size: 117.6 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
fc87c7f2579261539cac75adba908770a6e022e69041c64db20828b1bc1277f3
|
|
| MD5 |
f237afc9509925c3bc910749353f3238
|
|
| BLAKE2b-256 |
98b9316a66ebf788eb39f8418cb4ec2bd95f94cd752459cd258a596d3a4cb973
|