diff --git a/src/main.py b/src/main.py
new file mode 100644
index 0000000..9bce353
--- /dev/null
+++ b/src/main.py
@@ -0,0 +1,721 @@
+"""CLI entry point for ai-chat-exporter."""
+
+import logging
+import platform
+import shutil
+import sys
+import traceback
+from datetime import datetime
+from pathlib import Path
+
+import click
+from rich.console import Console
+from rich.table import Table
+
+from src.cache import Cache, CacheError
+from src.config import ConfigError
+from src.logging_config import setup_logging
+from src.providers.base import ProviderError
+
+console = Console()
+err_console = Console(stderr=True)
+
+TOS_NOTICE = """\
+⚠️  IMPORTANT — TERMS OF SERVICE NOTICE
+
+This tool accesses ChatGPT and Claude using internal web API endpoints
+that are not officially supported or documented by OpenAI or Anthropic.
+
+This approach may conflict with their Terms of Service:
+  • OpenAI ToS:     https://openai.com/policies/terms-of-use
+  • Anthropic ToS:  https://www.anthropic.com/legal/consumer-terms
+
+By using this tool you accept that:
+  - You are using it at your own risk
+  - Your account could potentially be suspended for automated access
+  - These APIs may break at any time without notice
+  - This tool is for personal archival use only, not commercial use
+
+Type 'yes' to acknowledge and continue, or Ctrl+C to exit: \
+"""
+
+# ──────────────────────────────────────────────────────────────────────────────
+# CLI group
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+@click.group()
+@click.version_option(version="0.1.0", prog_name="ai-chat-exporter")
+@click.option("--verbose", "-v", is_flag=True, help="Enable DEBUG output to console.")
+@click.option("--quiet", "-q", is_flag=True, help="Show WARNING and above only.")
+@click.option("--debug", is_flag=True, help="DEBUG + full tracebacks + redacted API bodies.")
+@click.option("--no-log-file", is_flag=True, help="Disable file logging.")
+@click.pass_context
+def cli(ctx: click.Context, verbose: bool, quiet: bool, debug: bool, no_log_file: bool) -> None:
+    """Export ChatGPT and Claude conversations to Markdown for personal archival."""
+    ctx.ensure_object(dict)
+
+    # Determine console log level
+    if debug or verbose:
+        level = logging.DEBUG
+    elif quiet:
+        level = logging.WARNING
+    else:
+        level = logging.INFO
+
+    # Determine log file path from env (setup_logging handles "none")
+    import os
+    log_file = os.getenv("LOG_FILE", "~/.ai-chat-exporter/logs/exporter.log")
+
+    setup_logging(level=level, log_file=log_file, no_log_file=no_log_file)
+
+    ctx.obj["debug"] = debug
+    ctx.obj["verbose"] = verbose
+
+    # Initialise cache (needed for ToS gate on every command)
+    import os
+    cache_dir = Path(os.getenv("CACHE_DIR", "~/.ai-chat-exporter")).expanduser()
+    try:
+        cache = Cache(cache_dir)
+    except CacheError as e:
+        err_console.print(f"[red]Cache error: {e}[/red]")
+        sys.exit(1)
+
+    ctx.obj["cache"] = cache
+
+    # ToS gate: must happen before any command executes
+    if not cache.is_tos_acknowledged():
+        try:
+            answer = click.prompt(TOS_NOTICE, default="", show_default=False).strip().lower()
+        except (click.Abort, KeyboardInterrupt):
+            console.print("\n[yellow]Exiting.[/yellow]")
+            sys.exit(0)
+        if answer != "yes":
+            console.print("[yellow]You must type 'yes' to proceed. Exiting.[/yellow]")
+            sys.exit(0)
+        cache.acknowledge_tos()
+        console.print("[green]Acknowledged. Proceeding.[/green]\n")
+
+
+# ──────────────────────────────────────────────────────────────────────────────
+# auth command
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+@cli.command()
+@click.pass_context
+def auth(ctx: click.Context) -> None:
+    """Interactive setup wizard for session tokens.
+
+    Guides you through finding and saving your ChatGPT and Claude session
+    tokens. Tokens are never echoed to the terminal.
+
+    Token lifetimes:
+      ChatGPT (__Secure-next-auth.session-token): ~7 days (JWT)
+      Claude (sessionKey): ~30 days (opaque string)
+    """
+    os_name = platform.system()
+
+    console.print("\n[bold cyan]AI Chat Exporter — Token Setup Wizard[/bold cyan]\n")
+    console.print("Session tokens let this tool access your chat history without your password.")
+    console.print("They are stored in browser cookies and have limited lifetimes.\n")
+
+    # Offer choice of provider to configure
+    providers = click.prompt(
+        "Which provider(s) do you want to configure? [chatgpt/claude/both]",
+        default="both",
+    ).strip().lower()
+
+    configure_chatgpt = providers in ("chatgpt", "both")
+    configure_claude = providers in ("claude", "both")
+
+    if configure_chatgpt:
+        _auth_chatgpt(os_name)
+
+    if configure_claude:
+        _auth_claude(os_name)
+
+    console.print("\n[green]Done! Run 'python -m src.main doctor' to verify your setup.[/green]")
+
+
+def _auth_chatgpt(os_name: str) -> None:
+    import jwt as pyjwt
+
+    console.print("\n[bold]─── ChatGPT ───[/bold]")
+    console.print("1. Open [link=https://chatgpt.com]https://chatgpt.com[/link] and log in.")
+    if os_name == "Darwin":
+        console.print("2. Press [bold]Cmd+Option+I[/bold] to open DevTools → Application tab.")
+    else:
+        console.print("2. Press [bold]F12[/bold] to open DevTools → Application tab.")
+    console.print("3. Expand [bold]Cookies[/bold] → [bold]https://chatgpt.com[/bold]")
+    console.print("4. Find [bold]__Secure-next-auth.session-token[/bold] → copy the Value.")
+    console.print("   (Token starts with 'eyJ...' — it is a long JWT string)")
+    console.print("5. Paste it below (input is hidden).\n")
+
+    token = click.prompt("ChatGPT session token", hide_input=True, default="", show_default=False).strip()
+    if not token:
+        console.print("[yellow]Skipped ChatGPT token.[/yellow]")
+        return
+
+    # Validate
+    if not token.startswith("eyJ"):
+        console.print("[yellow]Warning: token doesn't look like a JWT (expected 'eyJ...').[/yellow]")
+
+    expiry_str = ""
+    try:
+        payload = pyjwt.decode(token, options={"verify_signature": False})
+        exp = payload.get("exp")
+        if exp:
+            from datetime import timezone
+            expiry = datetime.fromtimestamp(exp, tz=timezone.utc)
+            expiry_str = expiry.strftime("%Y-%m-%d %H:%M UTC")
+            console.print(f"[green]Token decoded — expires: {expiry_str}[/green]")
+    except Exception:
+        console.print("[yellow]Could not decode token expiry.[/yellow]")
+
+    _write_token_to_env("CHATGPT_SESSION_TOKEN", token)
+
+
+def _auth_claude(os_name: str) -> None:
+    console.print("\n[bold]─── Claude ───[/bold]")
+    console.print("1. Open [link=https://claude.ai]https://claude.ai[/link] and log in.")
+    if os_name == "Darwin":
+        console.print("2. Press [bold]Cmd+Option+I[/bold] to open DevTools → Application tab.")
+    else:
+        console.print("2. Press [bold]F12[/bold] to open DevTools → Application tab.")
+    console.print("3. Expand [bold]Cookies[/bold] → [bold]https://claude.ai[/bold]")
+    console.print("4. Find [bold]sessionKey[/bold] → copy the Value.")
+    console.print("   (Note: Claude tokens expire after ~30 days; a 401 error is the only signal.)")
+    console.print("5. Paste it below (input is hidden).\n")
+
+    key = click.prompt("Claude session key", hide_input=True, default="", show_default=False).strip()
+    if not key:
+        console.print("[yellow]Skipped Claude token.[/yellow]")
+        return
+
+    console.print("[green]Claude session key saved.[/green]")
+    _write_token_to_env("CLAUDE_SESSION_KEY", key)
+
+
+def _write_token_to_env(key: str, value: str) -> None:
+    """Write or update a key in .env, offering to create the file if it doesn't exist."""
+    env_path = Path(".env")
+
+    if click.confirm(f"Write {key} to .env?", default=True):
+        if not env_path.exists():
+            # Create from example if available
+            example = Path(".env.example")
+            if example.exists():
+                import shutil as _shutil
+                _shutil.copy2(example, env_path)
+                console.print("[dim]Created .env from .env.example[/dim]")
+            else:
+                env_path.touch()
+
+        lines = env_path.read_text(encoding="utf-8").splitlines(keepends=True)
+        updated = False
+        new_lines = []
+        for line in lines:
+            if line.startswith(f"{key}=") or line.startswith(f"{key} ="):
+                new_lines.append(f"{key}={value}\n")
+                updated = True
+            else:
+                new_lines.append(line)
+
+        if not updated:
+            new_lines.append(f"\n{key}={value}\n")
+
+        env_path.write_text("".join(new_lines), encoding="utf-8")
+        import os
+        os.chmod(env_path, 0o600)
+        console.print(f"[green]{key} written to .env (permissions: 600)[/green]")
+
+
+# ──────────────────────────────────────────────────────────────────────────────
+# doctor command
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+@cli.command()
+@click.pass_context
+def doctor(ctx: click.Context) -> None:
+    """Validate configuration and API connectivity.
+
+    Checks token presence, format, expiry, directory permissions, disk space,
+    and live API reachability. Exits with code 1 if any checks fail.
+    """
+    checks = _run_doctor_checks()
+    _print_doctor_table(checks)
+
+    if any(not c["pass"] for c in checks):
+        sys.exit(1)
+
+
+def _run_doctor_checks() -> list[dict]:
+    """Run all doctor checks and return results."""
+    import os
+    import jwt as pyjwt
+    from datetime import timezone
+
+    checks = []
+
+    def add(name: str, passed: bool, detail: str = "") -> None:
+        checks.append({"name": name, "pass": passed, "detail": detail})
+
+    # Token presence
+    chatgpt_token = os.getenv("CHATGPT_SESSION_TOKEN", "").strip() or None
+    claude_key = os.getenv("CLAUDE_SESSION_KEY", "").strip() or None
+
+    add("CHATGPT_SESSION_TOKEN set", bool(chatgpt_token), "" if chatgpt_token else "Set in .env or run 'auth'")
+    add("CLAUDE_SESSION_KEY set", bool(claude_key), "" if claude_key else "Set in .env or run 'auth'")
+
+    # ChatGPT token format + expiry
+    if chatgpt_token:
+        is_jwt = chatgpt_token.startswith("eyJ")
+        add("ChatGPT token is valid JWT", is_jwt, "" if is_jwt else "Expected token starting with 'eyJ'")
+        if is_jwt:
+            try:
+                payload = pyjwt.decode(chatgpt_token, options={"verify_signature": False})
+                exp = payload.get("exp")
+                if exp:
+                    expiry = datetime.fromtimestamp(exp, tz=timezone.utc)
+                    now = datetime.now(tz=timezone.utc)
+                    delta = expiry - now
+                    detail = f"Expires {expiry.strftime('%Y-%m-%d %H:%M UTC')} ({delta.days}d)"
+                    ok = delta.total_seconds() > 0
+                    add("ChatGPT token not expired", ok, detail)
+                    if ok and delta.total_seconds() < 86400:
+                        add("ChatGPT token expiry warning", False, "Expires in < 24h — refresh soon")
+                else:
+                    add("ChatGPT token expiry", False, "JWT has no 'exp' claim")
+            except Exception as e:
+                add("ChatGPT token decode", False, str(e))
+
+    # Claude key
+    if claude_key:
+        add("Claude session key non-empty", True, "Expiry cannot be decoded (opaque token)")
+
+    # Directories
+    export_dir = Path(os.getenv("EXPORT_DIR", "./exports")).expanduser()
+    cache_dir = Path(os.getenv("CACHE_DIR", "~/.ai-chat-exporter")).expanduser()
+
+    for label, dirpath in [("Export dir writable", export_dir), ("Cache dir writable", cache_dir)]:
+        try:
+            dirpath.mkdir(parents=True, exist_ok=True)
+            test = dirpath / ".doctor_write_test"
+            test.touch()
+            test.unlink()
+            add(label, True, str(dirpath))
+        except OSError as e:
+            add(label, False, str(e))
+
+    # Disk space
+    try:
+        usage = shutil.disk_usage(export_dir if export_dir.exists() else Path("."))
+        free_mb = usage.free // (1024 * 1024)
+        add("Disk space (≥100MB free)", free_mb >= 100, f"{free_mb}MB free")
+    except OSError as e:
+        add("Disk space check", False, str(e))
+
+    # API reachability
+    if chatgpt_token:
+        try:
+            from src.providers.chatgpt import ChatGPTProvider
+            p = ChatGPTProvider(chatgpt_token)
+            results = p.list_conversations(offset=0, limit=1)
+            add("ChatGPT API reachable", True, f"Got {len(results)} result(s)")
+        except ProviderError as e:
+            add("ChatGPT API reachable", False, str(e.original)[:80])
+        except Exception as e:
+            add("ChatGPT API reachable", False, str(e)[:80])
+
+    if claude_key:
+        try:
+            from src.providers.claude import ClaudeProvider
+            p = ClaudeProvider(claude_key)
+            results = p.list_conversations(offset=0, limit=1)
+            add("Claude API reachable", True, f"Got {len(results)} result(s)")
+        except ProviderError as e:
+            add("Claude API reachable", False, str(e.original)[:80])
+        except Exception as e:
+            add("Claude API reachable", False, str(e)[:80])
+
+    return checks
+
+
+def _print_doctor_table(checks: list[dict]) -> None:
+    table = Table(title="Doctor Check Results", show_header=True)
+    table.add_column("Check", style="bold")
+    table.add_column("Status", justify="center")
+    table.add_column("Detail")
+
+    for c in checks:
+        status = "[green]✓ PASS[/green]" if c["pass"] else "[red]✗ FAIL[/red]"
+        table.add_row(c["name"], status, c.get("detail", ""))
+
+    console.print(table)
+
+
+# ──────────────────────────────────────────────────────────────────────────────
+# export command
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+@cli.command()
+@click.option(
+    "--provider",
+    type=click.Choice(["chatgpt", "claude", "all"], case_sensitive=False),
+    default="all",
+    show_default=True,
+    help="Which provider to export.",
+)
+@click.option(
+    "--format",
+    "fmt",
+    type=click.Choice(["markdown", "json", "both"], case_sensitive=False),
+    default="markdown",
+    show_default=True,
+    help="Output format.",
+)
+@click.option(
+    "--output",
+    "output_dir",
+    default=None,
+    type=click.Path(),
+    help="Override EXPORT_DIR.",
+)
+@click.option(
+    "--since",
+    default=None,
+    help="Only export conversations updated after this date (YYYY-MM-DD).",
+)
+@click.option("--dry-run", is_flag=True, help="Show what would be exported without writing anything.")
+@click.pass_context
+def export(
+    ctx: click.Context,
+    provider: str,
+    fmt: str,
+    output_dir: str | None,
+    since: str | None,
+    dry_run: bool,
+) -> None:
+    """Export new and updated conversations to Markdown or JSON.
+
+    Every run is automatically resumable — only new or updated conversations
+    are exported. Re-running the same command after an interruption will pick
+    up exactly where it left off.
+    """
+    debug = ctx.obj.get("debug", False)
+    cache: Cache = ctx.obj["cache"]
+
+    # Load config (may raise ConfigError)
+    try:
+        from src.config import load_config
+        cfg = _load_config_or_exit(debug)
+    except SystemExit:
+        return
+
+    # First-run: auto-doctor
+    if not cache.last_run():
+        console.print("[dim]First run — checking configuration…[/dim]")
+        checks = _run_doctor_checks()
+        _print_doctor_table(checks)
+        if any(not c["pass"] for c in checks):
+            err_console.print(
+                "[red]Doctor checks failed. Fix the issues above before exporting.[/red]"
+            )
+            sys.exit(1)
+
+    export_base = Path(output_dir).expanduser() if output_dir else cfg.export_dir
+    structure = cfg.output_structure
+
+    # Resolve since datetime
+    since_dt: datetime | None = None
+    if since:
+        try:
+            since_dt = datetime.fromisoformat(since)
+        except ValueError:
+            err_console.print(f"[red]Invalid --since date: '{since}'. Use YYYY-MM-DD.[/red]")
+            sys.exit(1)
+
+    # Determine which providers to run
+    providers_to_run = _resolve_providers(provider, cfg)
+    if not providers_to_run:
+        err_console.print(
+            "[red]No providers configured. Run 'python -m src.main auth' to set up tokens.[/red]"
+        )
+        sys.exit(1)
+
+    # Build exporters
+    from src.exporters.markdown import MarkdownExporter
+    from src.exporters.json_export import JSONExporter
+
+    md_exporter = MarkdownExporter(export_base, structure) if fmt in ("markdown", "both") else None
+    json_exporter = JSONExporter(export_base, structure) if fmt in ("json", "both") else None
+
+    # Summary counters
+    summary: dict[str, dict[str, int]] = {}
+
+    for prov_name, prov_instance in providers_to_run:
+        summary[prov_name] = {"exported": 0, "skipped": 0, "failed": 0}
+
+        console.print(f"\n[bold cyan][{prov_name.upper()}][/bold cyan] Fetching conversation list…")
+
+        try:
+            all_convs = prov_instance.fetch_all_conversations(since=since_dt)
+        except ProviderError as e:
+            _handle_provider_error(e, debug)
+            summary[prov_name]["failed"] += len(all_convs) if "all_convs" in dir() else 0
+            continue
+
+        to_export = cache.get_new_or_updated(prov_name, all_convs)
+        skipped = len(all_convs) - len(to_export)
+        summary[prov_name]["skipped"] = skipped
+
+        if dry_run:
+            _print_dry_run_table(prov_name, to_export, prov_instance, export_base, structure, skipped)
+            continue
+
+        if not to_export:
+            console.print(f"  [dim]{skipped} conversations already up to date.[/dim]")
+            continue
+
+        console.print(f"  [dim]{len(to_export)} to export, {skipped} already up to date.[/dim]")
+
+        from rich.progress import Progress, SpinnerColumn, TextColumn, BarColumn, TaskProgressColumn
+
+        with Progress(
+            SpinnerColumn(),
+            TextColumn("[progress.description]{task.description}"),
+            BarColumn(),
+            TaskProgressColumn(),
+            console=console,
+        ) as progress:
+            task = progress.add_task(f"Exporting {prov_name}…", total=len(to_export))
+
+            for raw_conv in to_export:
+                conv_id = raw_conv.get("id") or raw_conv.get("uuid", "unknown")
+                try:
+                    full_raw = prov_instance.get_conversation(conv_id)
+                    normalized = prov_instance.normalize_conversation(full_raw)
+
+                    exported_path: Path | None = None
+                    if md_exporter:
+                        exported_path = md_exporter.export(normalized)
+                    if json_exporter:
+                        exported_path = json_exporter.export(normalized)
+
+                    # Write to cache immediately after successful export
+                    cache.mark_exported(prov_name, conv_id, {
+                        "title": normalized.get("title", ""),
+                        "project": normalized.get("project"),
+                        "updated_at": normalized.get("updated_at", ""),
+                        "file_path": str(exported_path) if exported_path else "",
+                    })
+                    summary[prov_name]["exported"] += 1
+                    progress.advance(task)
+
+                except ProviderError as e:
+                    logger = logging.getLogger(__name__)
+                    logger.error("Failed to export conversation %s: %s", conv_id[:8], e)
+                    summary[prov_name]["failed"] += 1
+                    progress.advance(task)
+                    continue
+                except OSError as e:
+                    logger = logging.getLogger(__name__)
+                    logger.error("File write failed for conversation %s: %s", conv_id[:8], e)
+                    summary[prov_name]["failed"] += 1
+                    progress.advance(task)
+                    continue
+
+    if not dry_run:
+        _print_export_summary(summary)
+
+
+def _resolve_providers(provider: str, cfg) -> list[tuple[str, object]]:
+    """Return (name, instance) pairs for configured providers."""
+    result = []
+
+    def try_add(prov_name: str, token: str | None, provider_cls):
+        if not token:
+            if provider == prov_name or provider == "all":
+                logging.getLogger(__name__).warning(
+                    "[%s] Skipping — token not configured.", prov_name
+                )
+            return
+        try:
+            result.append((prov_name, provider_cls(token)))
+        except ProviderError as e:
+            logging.getLogger(__name__).warning(
+                "[%s] Could not initialise provider: %s", prov_name, e
+            )
+
+    from src.providers.chatgpt import ChatGPTProvider
+    from src.providers.claude import ClaudeProvider
+
+    if provider in ("chatgpt", "all"):
+        try_add("chatgpt", cfg.chatgpt_session_token, ChatGPTProvider)
+    if provider in ("claude", "all"):
+        try_add("claude", cfg.claude_session_key, ClaudeProvider)
+
+    return result
+
+
+def _print_dry_run_table(prov_name, to_export, prov_instance, export_base, structure, skipped):
+    from src.utils import generate_filename, build_export_path
+
+    table = Table(title=f"[DRY RUN] {prov_name.upper()} — Would export {len(to_export)} conversations")
+    table.add_column("Title")
+    table.add_column("Project")
+    table.add_column("Destination")
+    table.add_column("Updated")
+
+    for conv in to_export[:50]:  # cap display at 50
+        title = conv.get("title") or "Untitled"
+        project = conv.get("project_title") or conv.get("project", {})
+        if isinstance(project, dict):
+            project = project.get("name", "")
+        project = project or None
+        updated = (conv.get("updated_at") or conv.get("update_time") or "")[:10]
+        created = conv.get("created_at") or conv.get("create_time") or ""
+        conv_id = conv.get("id") or conv.get("uuid", "")
+        filename = generate_filename(title, conv_id, created or "2000-01-01")
+        dest = build_export_path(export_base, prov_name, project, created or "2000-01-01", filename, structure)
+        table.add_row(title[:50], str(project or "no-project"), str(dest), updated)
+
+    if len(to_export) > 50:
+        table.add_row(f"… and {len(to_export) - 50} more", "", "", "")
+
+    console.print(table)
+    console.print(f"  [dim]{skipped} conversations already cached (would be skipped).[/dim]")
+
+
+def _print_export_summary(summary: dict[str, dict[str, int]]) -> None:
+    table = Table(title="Export Summary")
+    table.add_column("Provider", style="bold")
+    table.add_column("Exported", justify="right")
+    table.add_column("Skipped", justify="right")
+    table.add_column("Failed", justify="right")
+
+    for prov, counts in summary.items():
+        table.add_row(
+            prov.capitalize(),
+            str(counts["exported"]),
+            str(counts["skipped"]),
+            f"[red]{counts['failed']}[/red]" if counts["failed"] else "0",
+        )
+
+    console.print(table)
+
+
+# ──────────────────────────────────────────────────────────────────────────────
+# list command
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+@cli.command(name="list")
+@click.option(
+    "--provider",
+    type=click.Choice(["chatgpt", "claude", "all"], case_sensitive=False),
+    default="all",
+    show_default=True,
+)
+@click.pass_context
+def list_conversations(ctx: click.Context, provider: str) -> None:
+    """List conversations without exporting them."""
+    debug = ctx.obj.get("debug", False)
+    cfg = _load_config_or_exit(debug)
+    providers_to_run = _resolve_providers(provider, cfg)
+
+    for prov_name, prov_instance in providers_to_run:
+        console.print(f"\n[bold cyan][{prov_name.upper()}][/bold cyan]")
+        try:
+            all_convs = prov_instance.fetch_all_conversations()
+        except ProviderError as e:
+            _handle_provider_error(e, debug)
+            continue
+
+        table = Table()
+        table.add_column("Title")
+        table.add_column("Project")
+        table.add_column("Updated")
+        table.add_column("ID")
+
+        for conv in all_convs:
+            title = conv.get("title") or "Untitled"
+            project = conv.get("project_title") or ""
+            if isinstance(conv.get("project"), dict):
+                project = conv["project"].get("name", "")
+            updated = (conv.get("updated_at") or conv.get("update_time") or "")[:10]
+            conv_id = (conv.get("id") or conv.get("uuid") or "")[:8]
+            table.add_row(title[:60], project[:30], updated, conv_id)
+
+        console.print(table)
+        console.print(f"Total: {len(all_convs)} conversations")
+
+
+# ──────────────────────────────────────────────────────────────────────────────
+# cache command
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+@cli.command()
+@click.option("--show", is_flag=True, help="Show cache statistics.")
+@click.option("--clear", is_flag=True, help="Clear cached entries.")
+@click.option(
+    "--provider",
+    type=click.Choice(["chatgpt", "claude", "all"], case_sensitive=False),
+    default="all",
+    help="Provider to target (used with --clear).",
+)
+@click.pass_context
+def cache(ctx: click.Context, show: bool, clear: bool, provider: str) -> None:
+    """Manage the local export cache (sync manifest)."""
+    cache_obj: Cache = ctx.obj["cache"]
+
+    if show:
+        stats = cache_obj.stats()
+        table = Table(title="Cache Statistics")
+        table.add_column("Provider", style="bold")
+        table.add_column("Cached Conversations", justify="right")
+        for prov, count in stats.items():
+            table.add_row(prov.capitalize(), str(count))
+        last = cache_obj.last_run()
+        console.print(table)
+        console.print(f"Last run: {last or 'never'}")
+
+    if clear:
+        prov_arg = None if provider == "all" else provider
+        if click.confirm(f"Clear cache for {'all providers' if prov_arg is None else prov_arg}?"):
+            cache_obj.clear(prov_arg)
+            console.print("[green]Cache cleared.[/green]")
+
+    if not show and not clear:
+        console.print("Specify --show or --clear. Use --help for options.")
+
+
+# ──────────────────────────────────────────────────────────────────────────────
+# Helpers
+# ──────────────────────────────────────────────────────────────────────────────
+
+
+def _load_config_or_exit(debug: bool):
+    """Load config, printing a clean error on failure."""
+    from src.config import load_config
+    try:
+        return load_config()
+    except ConfigError as e:
+        err_console.print(f"[red]Configuration error:[/red] {e}")
+        if debug:
+            traceback.print_exc()
+        sys.exit(1)
+
+
+def _handle_provider_error(e: ProviderError, debug: bool) -> None:
+    err_console.print(f"[red]Provider error ({e.provider_name}/{e.operation}):[/red] {e.original}")
+    if debug:
+        traceback.print_exc()
+
+
+if __name__ == "__main__":
+    cli()