diff --git a/veetcode/app.py b/veetcode/app.py
index dcf2c80..cf821e1 100644
--- a/veetcode/app.py
+++ b/veetcode/app.py
@@ -1 +1,433 @@
-# Placeholder
+"""Veetcode TUI - Terminal-based LeetCode practice with auto-testing."""
+
+from __future__ import annotations
+
+import json
+import re
+import subprocess
+import sys
+import time
+from dataclasses import dataclass, field
+from pathlib import Path
+
+from textual import work
+from textual.app import App, ComposeResult
+from textual.binding import Binding
+from textual.containers import Horizontal, Vertical, VerticalScroll
+from textual.screen import Screen
+from textual.widgets import Footer, Header, OptionList, RichLog, Static
+from textual.widgets.option_list import Option
+
+
+# ─────────────────────────────────────────────────────────────────────────────
+# Data
+# ─────────────────────────────────────────────────────────────────────────────
+
+
+@dataclass(frozen=True)
+class Problem:
+    name: str
+    difficulty: str
+    path: Path
+
+    @property
+    def solution_file(self) -> Path:
+        return self.path / "solution.py"
+
+
+@dataclass
+class TestCase:
+    name: str
+    passed: bool
+    error: str = ""
+
+
+@dataclass
+class SolutionStats:
+    lines: int = 0
+    code_lines: int = 0
+    functions: int = 0
+    complexity_hint: str = ""
+
+
+@dataclass
+class TestResult:
+    passed: bool
+    output: str
+    test_cases: list[TestCase] = field(default_factory=list)
+    total: int = 0
+    failed_count: int = 0
+    passed_count: int = 0
+    execution_time_ms: float = 0.0
+
+
+def strip_ansi(text: str) -> str:
+    """Remove ANSI escape codes from text."""
+    return re.sub(r"\x1b\[[0-9;]*m", "", text)
+
+
+def find_repo_root(start: Path | None = None) -> Path:
+    cur = (start or Path.cwd()).resolve()
+    for p in (cur, *cur.parents):
+        if (p / "pyproject.toml").exists():
+            return p
+    return cur
+
+
+def load_solved(path: Path) -> set[str]:
+    if path.exists():
+        try:
+            return set(json.loads(path.read_text()))
+        except (json.JSONDecodeError, TypeError):
+            pass
+    return set()
+
+
+def save_solved(path: Path, solved: set[str]) -> None:
+    path.write_text(json.dumps(sorted(solved)))
+
+
+def scan_problems(problems_dir: Path) -> list[Problem]:
+    problems: list[Problem] = []
+    for difficulty in ("easy", "medium", "hard"):
+        diff_dir = problems_dir / difficulty
+        if not diff_dir.exists():
+            continue
+        for problem_dir in sorted(diff_dir.iterdir()):
+            if problem_dir.is_dir() and (problem_dir / "solution.py").exists():
+                problems.append(
+                    Problem(name=problem_dir.name, difficulty=difficulty, path=problem_dir)
+                )
+    return problems
+
+
+def analyze_solution(solution_file: Path) -> SolutionStats:
+    stats = SolutionStats()
+    if not solution_file.exists():
+        return stats
+    try:
+        code = solution_file.read_text()
+        lines = code.split("\n")
+        stats.lines = len(lines)
+        for line in lines:
+            stripped = line.strip()
+            if stripped and not stripped.startswith("#"):
+                stats.code_lines += 1
+            if stripped.startswith("def "):
+                stats.functions += 1
+
+        # Simple complexity hints
+        hints = []
+        if "for " in code and "for " in code[code.find("for ") + 4:]:
+            hints.append("O(n²)")
+        elif "while " in code or "for " in code:
+            hints.append("O(n)")
+        if "sorted(" in code or ".sort(" in code:
+            hints.append("sort")
+        stats.complexity_hint = ", ".join(hints[:2]) if hints else "—"
+    except Exception:
+        pass
+    return stats
+
+
+def parse_pytest_output(output: str) -> tuple[list[TestCase], float]:
+    """Parse pytest output to extract test results."""
+    test_cases: list[TestCase] = []
+    total_time = 0.0
+
+    # Strip ANSI codes for reliable parsing
+    clean = strip_ansi(output)
+
+    # Match: tests.py::test_name PASSED or FAILED
+    for match in re.finditer(r"tests\.py::(\w+)\s+(PASSED|FAILED)", clean):
+        test_cases.append(TestCase(
+            name=match.group(1),
+            passed=match.group(2) == "PASSED"
+        ))
+
+    # Extract timing from summary like "4 passed in 0.12s"
+    time_match = re.search(r"in\s+([\d.]+)s", clean)
+    if time_match:
+        total_time = float(time_match.group(1)) * 1000
+
+    # Extract errors for failed tests
+    for match in re.finditer(r"FAILED tests\.py::(\w+)\s*-\s*(\w+:.*?)(?=\n|$)", clean):
+        test_name, error = match.group(1), match.group(2).strip()[:100]
+        for tc in test_cases:
+            if tc.name == test_name and not tc.passed:
+                tc.error = error
+                break
+
+    # Fallback: parse summary line for counts if no individual tests found
+    if not test_cases:
+        # Match "1 passed" or "2 failed" etc
+        passed_match = re.search(r"(\d+)\s+passed", clean)
+        failed_match = re.search(r"(\d+)\s+failed", clean)
+        if passed_match or failed_match:
+            passed_count = int(passed_match.group(1)) if passed_match else 0
+            failed_count = int(failed_match.group(1)) if failed_match else 0
+            for i in range(passed_count):
+                test_cases.append(TestCase(name=f"test_{i+1}", passed=True))
+            for i in range(failed_count):
+                test_cases.append(TestCase(name=f"test_{passed_count+i+1}", passed=False))
+
+    return test_cases, total_time
+
+
+def run_tests(problem_path: Path) -> TestResult:
+    tests_file = problem_path / "tests.py"
+    if not tests_file.exists():
+        return TestResult(passed=False, output="No tests.py found")
+
+    try:
+        start = time.perf_counter()
+        result = subprocess.run(
+            [sys.executable, "-m", "pytest", "tests.py", "-v", "--tb=short", "--no-header"],
+            cwd=problem_path,
+            capture_output=True,
+            text=True,
+            timeout=30,
+        )
+        elapsed = (time.perf_counter() - start) * 1000
+        output = result.stdout + result.stderr
+
+        test_cases, parsed_time = parse_pytest_output(output)
+        passed_count = sum(1 for tc in test_cases if tc.passed)
+        failed_count = len(test_cases) - passed_count
+
+        return TestResult(
+            passed=result.returncode == 0,
+            output=output,
+            test_cases=test_cases,
+            total=len(test_cases),
+            passed_count=passed_count,
+            failed_count=failed_count,
+            execution_time_ms=parsed_time or elapsed,
+        )
+    except subprocess.TimeoutExpired:
+        return TestResult(passed=False, output="Timeout (30s)")
+    except Exception as e:
+        return TestResult(passed=False, output=str(e))
+
+
+# ─────────────────────────────────────────────────────────────────────────────
+# Screens
+# ─────────────────────────────────────────────────────────────────────────────
+
+
+class ProblemListScreen(Screen):
+    BINDINGS = [
+        Binding("q", "quit", "Quit"),
+        Binding("r", "rescan", "Rescan"),
+        Binding("j", "cursor_down", show=False),
+        Binding("k", "cursor_up", show=False),
+    ]
+
+    def __init__(self) -> None:
+        super().__init__()
+        self.repo_root = find_repo_root()
+        self.problems_dir = self.repo_root / "problems"
+        self.solved_file = self.repo_root / ".solved.json"
+        self.problems: list[Problem] = []
+        self.solved: set[str] = set()
+
+    def compose(self) -> ComposeResult:
+        yield Header()
+        yield Vertical(
+            Static("veetcode", id="title"),
+            OptionList(id="problem-list"),
+            id="main-container",
+        )
+        yield Footer()
+
+    def on_mount(self) -> None:
+        self.load_problems()
+        self.query_one("#problem-list", OptionList).focus()
+
+    def load_problems(self) -> None:
+        self.problems = scan_problems(self.problems_dir)
+        self.solved = load_solved(self.solved_file)
+
+        ol = self.query_one("#problem-list", OptionList)
+        ol.clear_options()
+
+        if not self.problems:
+            ol.add_option(Option("[dim]No problems found[/dim]", disabled=True))
+            return
+
+        for p in self.problems:
+            diff_char = {"easy": "E", "medium": "M", "hard": "H"}.get(p.difficulty, "?")
+            check = "✓ " if p.name in self.solved else "  "
+            ol.add_option(Option(f"{check}[{diff_char}] {p.name}", id=p.name))
+
+    def action_rescan(self) -> None:
+        self.load_problems()
+        self.notify("Rescanned")
+
+    def action_quit(self) -> None:
+        self.app.exit()
+
+    def action_cursor_down(self) -> None:
+        self.query_one("#problem-list", OptionList).action_cursor_down()
+
+    def action_cursor_up(self) -> None:
+        self.query_one("#problem-list", OptionList).action_cursor_up()
+
+    def on_option_list_option_selected(self, event: OptionList.OptionSelected) -> None:
+        for p in self.problems:
+            if p.name == event.option.id:
+                self.app.push_screen(WatchScreen(p, self.solved_file, self.solved))
+                break
+
+
+class WatchScreen(Screen):
+    BINDINGS = [
+        Binding("escape", "back", "Back"),
+        Binding("r", "rerun", "Rerun"),
+        Binding("q", "quit", "Quit"),
+        Binding("j", "scroll_down", show=False),
+        Binding("k", "scroll_up", show=False),
+    ]
+
+    def __init__(self, problem: Problem, solved_file: Path, solved: set[str]) -> None:
+        super().__init__()
+        self.problem = problem
+        self.solved_file = solved_file
+        self.solved = solved
+        self.watcher = None
+
+    def compose(self) -> ComposeResult:
+        yield Header()
+        yield Vertical(
+            Static(f"{self.problem.name} ({self.problem.difficulty})", id="problem-title"),
+            Horizontal(
+                Vertical(
+                    Static("", id="summary-status"),
+                    VerticalScroll(Static("", id="test-summary"), id="summary-scroll"),
+                    id="summary-pane",
+                ),
+                Vertical(
+                    Static("", id="stats-content"),
+                    id="stats-pane",
+                ),
+                id="info-row",
+            ),
+            RichLog(id="output", highlight=True, markup=True),
+            id="watch-container",
+        )
+        yield Footer()
+
+    def on_mount(self) -> None:
+        self._update_stats()
+        self.run_tests_async()
+        self.start_watcher()
+        self.query_one("#output", RichLog).focus()
+
+    def on_unmount(self) -> None:
+        self.stop_watcher()
+
+    def _update_stats(self) -> None:
+        stats = analyze_solution(self.problem.solution_file)
+        self.query_one("#stats-content", Static).update(
+            f"Lines: {stats.lines} ({stats.code_lines} code)\n"
+            f"Functions: {stats.functions}\n"
+            f"Complexity: {stats.complexity_hint}"
+        )
+
+    def start_watcher(self) -> None:
+        from watchdog.events import FileSystemEventHandler
+        from watchdog.observers import Observer
+
+        screen = self
+
+        class Handler(FileSystemEventHandler):
+            def on_modified(self, event):
+                if event.src_path.endswith("solution.py"):
+                    screen.app.call_from_thread(screen._on_file_change)
+
+        self.watcher = Observer()
+        self.watcher.schedule(Handler(), str(self.problem.path), recursive=False)
+        self.watcher.start()
+
+    def _on_file_change(self) -> None:
+        self._update_stats()
+        self.run_tests_async()
+
+    def stop_watcher(self) -> None:
+        if self.watcher:
+            self.watcher.stop()
+            self.watcher.join(timeout=1)
+            self.watcher = None
+
+    @work(thread=True, exclusive=True)
+    def run_tests_async(self) -> None:
+        self.app.call_from_thread(
+            lambda: self.query_one("#summary-status", Static).update("Running...")
+        )
+        result = run_tests(self.problem.path)
+        self.app.call_from_thread(self._display_result, result)
+
+    def _display_result(self, result: TestResult) -> None:
+        # Status line
+        time_str = f"{result.execution_time_ms:.0f}ms"
+        if result.passed:
+            status = f"✓ {result.total} passed ({time_str})"
+            self.solved.add(self.problem.name)
+            save_solved(self.solved_file, self.solved)
+        else:
+            status = f"✗ {result.failed_count}/{result.total} failed ({time_str})"
+        self.query_one("#summary-status", Static).update(status)
+
+        # Test list
+        lines = []
+        for i, tc in enumerate(result.test_cases, 1):
+            mark = "✓" if tc.passed else "✗"
+            lines.append(f"  {mark} {tc.name}")
+            if tc.error:
+                lines.append(f"    {tc.error[:80]}")
+        self.query_one("#test-summary", Static).update("\n".join(lines) or "No tests")
+
+        # Verbose output
+        out = self.query_one("#output", RichLog)
+        out.clear()
+        out.write(result.output or "No output")
+
+    def action_back(self) -> None:
+        self.stop_watcher()
+        self.app.pop_screen()
+
+    def action_rerun(self) -> None:
+        self._update_stats()
+        self.run_tests_async()
+
+    def action_quit(self) -> None:
+        self.stop_watcher()
+        self.app.exit()
+
+    def action_scroll_down(self) -> None:
+        self.query_one("#output", RichLog).scroll_down()
+
+    def action_scroll_up(self) -> None:
+        self.query_one("#output", RichLog).scroll_up()
+
+
+# ─────────────────────────────────────────────────────────────────────────────
+# App
+# ─────────────────────────────────────────────────────────────────────────────
+
+
+class VeetcodeApp(App):
+    CSS_PATH = "veetcode.tcss"
+    TITLE = "veetcode"
+
+    def on_mount(self) -> None:
+        self.theme = "gruvbox"
+        self.push_screen(ProblemListScreen())
+
+
+def run_app() -> None:
+    VeetcodeApp().run()
+
+
+if __name__ == "__main__":
+    run_app()
diff --git a/veetcode/cli.py b/veetcode/cli.py
index dcf2c80..afc630c 100644
--- a/veetcode/cli.py
+++ b/veetcode/cli.py
@@ -1 +1,21 @@
-# Placeholder
+"""CLI entry point for veetcode."""
+
+import typer
+
+from veetcode.app import run_app
+
+app = typer.Typer(
+    name="veetcode",
+    help="Terminal-based LeetCode practice with auto-testing",
+    add_completion=False,
+)
+
+
+@app.command()
+def main() -> None:
+    """Launch the Veetcode TUI."""
+    run_app()
+
+
+if __name__ == "__main__":
+    app()
diff --git a/veetcode/veetcode.tcss b/veetcode/veetcode.tcss
new file mode 100644
index 0000000..ab2fb04
--- /dev/null
+++ b/veetcode/veetcode.tcss
@@ -0,0 +1,143 @@
+/* Veetcode TUI - Minimal, clean design */
+
+Screen {
+    background: $surface;
+}
+
+/* ─────────────────────────────────────────────────────────────────────────── */
+/* Problem List Screen */
+/* ─────────────────────────────────────────────────────────────────────────── */
+
+#main-container {
+    width: 100%;
+    height: 1fr;
+    padding: 1 2;
+}
+
+#title {
+    width: 100%;
+    height: auto;
+    text-align: center;
+    padding: 1;
+    color: $text-muted;
+}
+
+#problem-list {
+    width: 100%;
+    height: 1fr;
+    border: solid $primary-muted;
+    padding: 1;
+}
+
+OptionList {
+    scrollbar-gutter: stable;
+}
+
+OptionList:focus {
+    border: solid $accent;
+}
+
+OptionList > .option-list--option-highlighted {
+    background: $primary 20%;
+}
+
+/* ─────────────────────────────────────────────────────────────────────────── */
+/* Watch Screen */
+/* ─────────────────────────────────────────────────────────────────────────── */
+
+#watch-container {
+    width: 100%;
+    height: 1fr;
+    padding: 0 1;
+}
+
+#problem-title {
+    width: 100%;
+    height: auto;
+    text-align: center;
+    padding: 1;
+    color: $text;
+}
+
+/* Info Row */
+#info-row {
+    width: 100%;
+    height: auto;
+    min-height: 6;
+    max-height: 14;
+    margin: 0 0 1 0;
+}
+
+/* Summary Pane - Left */
+#summary-pane {
+    width: 2fr;
+    height: 100%;
+    border: solid $primary-muted;
+    padding: 0;
+}
+
+#summary-header {
+    display: none;
+}
+
+#summary-status {
+    width: 100%;
+    height: auto;
+    padding: 1;
+    text-align: left;
+}
+
+#summary-scroll {
+    width: 100%;
+    height: auto;
+    max-height: 8;
+    padding: 0 1;
+}
+
+#test-summary {
+    width: 100%;
+    height: auto;
+}
+
+/* Stats Pane - Right */
+#stats-pane {
+    width: 1fr;
+    height: 100%;
+    border: solid $primary-muted;
+    padding: 0;
+    margin-left: 1;
+}
+
+#stats-header {
+    display: none;
+}
+
+#stats-content {
+    width: 100%;
+    height: auto;
+    padding: 1;
+    color: $text-muted;
+}
+
+/* Verbose Output */
+#output-header {
+    display: none;
+}
+
+#output {
+    width: 100%;
+    height: 1fr;
+    min-height: 6;
+    border: solid $primary-muted;
+    padding: 1;
+    scrollbar-gutter: stable;
+}
+
+#output:focus {
+    border: solid $accent;
+}
+
+RichLog {
+    overflow-x: hidden;
+    text-wrap: wrap;
+}