diff --git a/EvoScientist/cli/tui_backends.py b/EvoScientist/cli/tui_backends.py
index ea51ce3..1d8bc87 100644
--- a/EvoScientist/cli/tui_backends.py
+++ b/EvoScientist/cli/tui_backends.py
@@ -34,7 +34,7 @@ class StreamingTUIBackend(Protocol):
 class RichStreamingBackend:
     """Default Rich backend wrapper around the existing streaming renderer."""
 
-    name: str = "rich"
+    name: str = "cli"
 
     def run_streaming(
         self,
diff --git a/EvoScientist/cli/tui_interactive.py b/EvoScientist/cli/tui_interactive.py
index 8f8b6a3..be07020 100644
--- a/EvoScientist/cli/tui_interactive.py
+++ b/EvoScientist/cli/tui_interactive.py
@@ -1290,7 +1290,7 @@ def run_textual_interactive(
                 memory_path = _shorten_path(str(paths.MEMORY_DIR))
                 if memory_path:
                     self._append_system(f"Memory dir: {memory_path}", style="dim")
-                self._append_system("UI: textual", style="dim")
+                self._append_system("UI: tui", style="dim")
                 return
 
             if cmd == "/new":
diff --git a/EvoScientist/config/onboard.py b/EvoScientist/config/onboard.py
index 3de1bb0..8184075 100644
--- a/EvoScientist/config/onboard.py
+++ b/EvoScientist/config/onboard.py
@@ -475,7 +475,7 @@ def _step_ui_backend(config: EvoScientistConfig) -> str:
         config: Current configuration.
 
     Returns:
-        Selected backend name ("rich" or "textual").
+        Selected backend name ("tui" or "cli").
     """
     choices = [
         Choice(title="TUI (full-screen interface, recommended)", value="tui"),
diff --git a/README.md b/README.md
index 9b91ab5..0eeb012 100644
--- a/README.md
+++ b/README.md
@@ -56,7 +56,7 @@ Going beyond traditional human-in-the-loop systems, EvoScientist introduces an A
 - **🔌 MCP & Skills** — Plug in MCP servers or install skills from GitHub on the fly.
 
 ## 🎯 ᯓ➤ Roadmap
-- [x] 🖥️ TUI powered by Rich and Textual
+- [x] 🖥️ Full-screen TUI and classic CLI interfaces
 - [x] 📻 EvoMemory v1.0 shipped
 - [x] ⚒️ 200+ predefined skills built in
 - [x] 🧩 Built-in research-lifecycle skills shipped
@@ -171,8 +171,8 @@ The easiest way to configure API keys is the interactive wizard:
 ```bash
 EvoSci onboard
 ```
-
-It walks you through provider selection, key validation, model choice, and workspace setup.
+> [!TIP]
+> It walks you through provider selection, key validation, model choice, and workspace mode.
 
 <details>
 <summary> 📟 Manual configuration via environment variables </summary>
@@ -207,7 +207,7 @@ cp .env.example .env  # then fill in your keys
 ## ⚡ Quick Start
 
 ```bash
-EvoSci  # or EvoScientist — interactive mode
+EvoSci  # or EvoScientist — interactive mode (TUI by default)
 ```
 
 ![demo](.github/assets/EvoScientist_cli.png)
@@ -220,9 +220,11 @@ EvoSci  # or EvoScientist — interactive mode
 <summary>Common examples</summary>
 
 ```bash
+EvoSci                            # interactive mode (TUI by default)
 EvoSci -p "your question"        # single-shot mode
+EvoSci --workdir /path/to/project # open in a specific directory
 EvoSci -m run                     # isolated per-session workspace
-EvoSci --ui textual               # alternative TUI backend
+EvoSci --ui cli                   # classic CLI (lightweight)
 EvoSci serve                      # headless mode — channels only, no interactive prompt
 ```
 
diff --git a/README.zh-CN.md b/README.zh-CN.md
index 2092782..826d723 100644
--- a/README.zh-CN.md
+++ b/README.zh-CN.md
@@ -59,13 +59,13 @@ EvoScientist 超越了传统的人在回路（Human-in-the-Loop）模式，引
 - **🤖 多智能体协作** — 6 个子智能体（规划、调研、编码、调试、分析、写作）协同工作。
 - **🧠 持久化记忆** — 上下文、偏好和研究发现跨会话保持。
 - **🔬 科学工作流** — 需求采集 → 规划 → 执行 → 评估 → 撰写 → 验证。
-- **🌐 多模型供应商** — Anthropic、OpenAI、Google、NVIDIA——一处配置，随时切换。
+- **🌐 多模型供应商** — Anthropic、OpenAI、Google、NVIDIA、SiliconFlow、OpenRouter、Ollama——一处配置，随时切换。
 - **📱 多渠道接入** — CLI 为中心；Telegram、Discord、Slack、飞书、微信等——共享同一智能体会话。
 - **🔌 MCP 与 Skills** — 即插即用 MCP 服务器，或从 GitHub 一键安装技能包。
 
 ## 🎯 ᯓ➤ 路线图
 
-- [x] 🖥️ 基于 Rich 和 Textual 的 TUI 界面
+- [x] 🖥️ 全屏 TUI 和经典 CLI 双界面
 - [x] 📻 EvoMemory v1.0 已上线
 - [x] ⚒️ 200+ 预置技能已内置
 - [x] 🧩 内置研究全流程技能已上线
@@ -182,7 +182,8 @@ git pull && uv sync --dev
 EvoSci onboard
 ```
 
-向导将引导你完成供应商选择、密钥验证、模型选择和工作区设置。
+> [!TIP]
+> 向导将引导你完成供应商选择、密钥验证、模型选择和工作区模式设置。
 
 <details>
 <summary> 📟 通过环境变量手动配置</summary>
@@ -217,7 +218,7 @@ cp .env.example .env  # 填入你的密钥
 ## ⚡ 快速上手
 
 ```bash
-EvoSci  # 或 EvoScientist — 交互模式
+EvoSci  # 或 EvoScientist — 交互模式（默认 TUI）
 ```
 
 ![demo](.github/assets/EvoScientist_cli.png)
@@ -230,9 +231,11 @@ EvoSci  # 或 EvoScientist — 交互模式
 <summary>常用示例</summary>
 
 ```bash
+EvoSci                            # 交互模式（默认 TUI）
 EvoSci -p "你的问题"              # 单次查询模式
+EvoSci --workdir /path/to/project # 在指定目录下启动
 EvoSci -m run                     # 隔离的会话级工作区
-EvoSci --ui textual               # 替代 TUI 后端
+EvoSci --ui cli                   # 经典 CLI（轻量）
 EvoSci serve                      # 无头模式——仅渠道，无交互提示符
 ```
 
diff --git a/tests/test_onboard.py b/tests/test_onboard.py
index 847e715..11d4c19 100644
--- a/tests/test_onboard.py
+++ b/tests/test_onboard.py
@@ -745,7 +745,7 @@ class TestRunOnboard:
 
             # Mock all questionary calls
             mock_q.select.return_value.ask.side_effect = [
-                "rich",  # UI backend
+                "tui",  # UI backend
                 "anthropic",  # Provider
                 "claude-sonnet-4-5",  # Model
                 "daemon",  # Workspace mode
@@ -796,7 +796,7 @@ class TestRunOnboard:
             mock_load.return_value = EvoScientistConfig()
 
             mock_q.select.return_value.ask.side_effect = [
-                "rich",  # UI backend
+                "tui",  # UI backend
                 "anthropic",
                 "claude-sonnet-4-5",
                 "daemon",