diff --git a/README.md b/README.md
index e7f27f0..0532746 100644
--- a/README.md
+++ b/README.md
@@ -56,43 +56,34 @@ phi login
 
 ## Quick start
 
-### Single-sequence / single-structure jobs
+### Try the tutorial
 
-```bash
-# Structure prediction (ESMFold)
-phi folding --fasta sequences.fasta
-
-# Complex structure prediction (AlphaFold2 multimer)
-phi complex_folding --fasta binder_target.fasta
+The fastest way to get started — downloads five example PD-L1 binder structures
+and walks you through the full pipeline:
 
-# Sequence design via inverse folding (ProteinMPNN)
-phi inverse_folding --pdb design.pdb --num-sequences 20
+```bash
+phi tutorial
 ```
 
-### Batch scoring workflow
+This fetches the example files, prints step-by-step instructions, and leaves
+you ready to run `phi filter`.
+
+### Scoring your own structures
 
 ```bash
-# 1. Upload a directory of PDB/CIF files
+# 1. Upload PDB/CIF files
 phi upload ./designs/
 
-# Output:
-#   dataset_id  d7c3a1b2-...
-#   Dashboard:  https://design.dynotx.com/dashboard/datasets/d7c3a1b2-...
-#   Run a job against this dataset:
-#     phi folding          --dataset-id d7c3a1b2-...
-#     phi complex_folding  --dataset-id d7c3a1b2-...
-#     phi inverse_folding  --dataset-id d7c3a1b2-...
-#     phi filter           --dataset-id d7c3a1b2-... --preset default --wait
-
-# 2. Run the full filter pipeline (inverse folding → folding → complex folding → score)
-phi filter --dataset-id d7c3a1b2-... --preset default --wait
+# 2. Run the full filter pipeline
+phi filter --preset default --wait
 
-# 3. Download results (structures, scores CSV, raw score JSONs)
+# 3. View scores and download results
+phi scores
 phi download --out ./results/
 ```
 
-After each command, `phi` prints the active dataset and job IDs and a link to
-the dashboard:
+After each command, `phi` prints the active dataset and a link to the
+dashboard:
 
 ```
 Active: dataset [d7c3a1b2-...] · job [cb4553f5-...]
@@ -105,6 +96,7 @@ Dashboard: https://design.dynotx.com/dashboard/datasets/d7c3a1b2-...
 
 | Command | Alias | Description |
 |---|---|---|
+| `phi tutorial` | — | Download example structures and print a step-by-step walkthrough |
 | `phi login` | — | Verify API key and print identity |
 | `phi upload` | — | Upload PDB/CIF files or a directory |
 | `phi fetch` | — | Download a structure from RCSB PDB or AlphaFold DB, crop, and optionally upload |
diff --git a/src/phi/cli.py b/src/phi/cli.py
index 2145e24..80d7440 100755
--- a/src/phi/cli.py
+++ b/src/phi/cli.py
@@ -16,6 +16,7 @@
 )
 from phi.commands.research import cmd_notes, cmd_research
 from phi.commands.structure import cmd_fetch
+from phi.commands.tutorial import cmd_tutorial
 from phi.config import _load_state
 from phi.display import _C_BLUE, _die, console
 from phi.parser import build_parser
@@ -23,6 +24,7 @@
 
 COMMANDS = {
     "login": cmd_login,
+    "tutorial": cmd_tutorial,
     "upload": cmd_upload,
     "ingest-session": cmd_ingest_session,
     "datasets": cmd_datasets,
diff --git a/src/phi/commands/models.py b/src/phi/commands/models.py
index 5233147..34be5b3 100644
--- a/src/phi/commands/models.py
+++ b/src/phi/commands/models.py
@@ -1,4 +1,7 @@
+from __future__ import annotations
+
 import argparse
+import os
 from pathlib import Path
 
 from phi.api import _require_key, _submit
@@ -6,6 +9,13 @@
 from phi.download import _download_job, _read_fasta
 from phi.polling import _poll
 
+_DESIGN_ENABLED = os.environ.get("DYNO_ENABLE_DESIGN", "").lower() in ("1", "true", "yes")
+
+
+def _require_design_flag() -> None:
+    if not _DESIGN_ENABLED:
+        _die("This command is not yet available.")
+
 
 def _run_model_job(job_type: str, params: dict, args: argparse.Namespace) -> None:
     from phi.config import POLL_INTERVAL as _INTERVAL
@@ -92,6 +102,7 @@ def cmd_boltz(args: argparse.Namespace) -> None:
 
 
 def cmd_rfdiffusion3(args: argparse.Namespace) -> None:
+    _require_design_flag()
     params: dict = {
         "num_designs": args.num_designs,
         "inference_steps": args.steps,
@@ -129,6 +140,7 @@ def cmd_rfdiffusion3(args: argparse.Namespace) -> None:
 
 
 def cmd_boltzgen(args: argparse.Namespace) -> None:
+    _require_design_flag()
     params: dict = {
         "protocol": args.protocol,
         "num_designs": args.num_designs,
diff --git a/src/phi/commands/research.py b/src/phi/commands/research.py
index c356c0f..7f0c47e 100644
--- a/src/phi/commands/research.py
+++ b/src/phi/commands/research.py
@@ -111,8 +111,8 @@ def cmd_notes(args: argparse.Namespace) -> None:
         return
 
     content: str = data.get("content") or ""
-    gcs_url: str | None = data.get("gcs_url")
-    gcs_uri: str | None = data.get("gcs_uri")
+    data.get("gcs_url")
+    data.get("gcs_uri")
 
     if args.out:
         out = Path(args.out)
@@ -124,8 +124,6 @@ def cmd_notes(args: argparse.Namespace) -> None:
         dest.parent.mkdir(parents=True, exist_ok=True)
         dest.write_text(content, encoding="utf-8")
         console.print(f"[{_C_SAND}]Notes saved[/] → {dest}")
-        if gcs_url:
-            console.print(f"[dim]Download URL:[/] {gcs_url}")
         return
 
     if args.json:
@@ -141,5 +139,3 @@ def cmd_notes(args: argparse.Namespace) -> None:
             padding=(1, 2),
         )
     )
-    if gcs_uri:
-        console.print(f"[dim]Storage URI:[/] {gcs_uri}")
diff --git a/src/phi/commands/structure.py b/src/phi/commands/structure.py
index aeca492..5685713 100644
--- a/src/phi/commands/structure.py
+++ b/src/phi/commands/structure.py
@@ -153,14 +153,12 @@ def _fetch_and_upload(
         ("Source  ", source_url),
         ("File    ", str(out_path)),
         ("Dataset ", dataset_id),
-        ("GCS URI ", gcs_uri),
     ]:
         console.print(f"  [bold]{label}[/bold] {value}")
     console.print()
     console.print("  [dim]Next steps:[/dim]")
-    console.print(
-        f"  [dim]  phi design --target-pdb-gcs {gcs_uri} --hotspots <A45,A67> --num-designs 50[/dim]"
-    )
+    console.print(f"  [dim]  phi upload {out_path}[/dim]")
+    console.print("  [dim]  phi filter --preset default --wait[/dim]")
     console.rule()
 
 
diff --git a/src/phi/commands/tutorial.py b/src/phi/commands/tutorial.py
new file mode 100644
index 0000000..2fd3f4b
--- /dev/null
+++ b/src/phi/commands/tutorial.py
@@ -0,0 +1,88 @@
+from __future__ import annotations
+
+import argparse
+import urllib.request
+from pathlib import Path
+
+from phi.api import _request
+from phi.config import _save_state, _ssl_context
+from phi.display import _C_BLUE, _C_SAND, _die, console
+
+
+def cmd_tutorial(args: argparse.Namespace) -> None:
+    out = Path(args.out)
+
+    # ── 1. Fetch manifest (standard Clerk JWT auth, same as all endpoints) ───
+    console.print("[dim]Fetching tutorial dataset …[/]")
+    try:
+        manifest = _request("GET", "/tutorial")
+    except Exception as exc:
+        _die(
+            f"Could not reach the tutorial endpoint: {exc}\n"
+            "  Check your connection and API key, then try again."
+        )
+
+    files: list[dict] = manifest.get("files", [])
+    dataset_id: str | None = manifest.get("dataset_id")
+    message: str | None = manifest.get("message")
+
+    if not files:
+        _die("No tutorial files returned by the API.")
+
+    # ── 2. Download each file (plain HTTP — signed URLs are self-authenticating)
+    out.mkdir(parents=True, exist_ok=True)
+    console.print(f"  Downloading {len(files)} file(s) to [{_C_BLUE}]{out}/[/] …\n")
+
+    for entry in files:
+        filename: str = entry["filename"]
+        url: str = entry["url"]
+        dest = out / filename
+        dest.parent.mkdir(parents=True, exist_ok=True)
+        try:
+            req = urllib.request.Request(url)
+            with urllib.request.urlopen(req, context=_ssl_context()) as resp:
+                dest.write_bytes(resp.read())
+            console.print(f"  [bold {_C_SAND}]✓[/]  {filename}")
+        except Exception as exc:
+            _die(f"Failed to download {filename}: {exc}")
+
+    # ── 3. Cache dataset_id so phi filter needs zero extra flags ─────────────
+    if dataset_id:
+        _save_state({"dataset_id": dataset_id})
+        console.print(
+            f"\n[dim]dataset_id [{_C_BLUE}]{dataset_id}[/] cached — "
+            f"run [bold]phi filter[/] to start scoring.[/]"
+        )
+
+    # ── 4. Print step-by-step guide ──────────────────────────────────────────
+    if message:
+        console.print(f"\n[dim]{message}[/]")
+
+    if dataset_id:
+        upload_step = "[dim]  (skipped — dataset already ready)[/]"
+    else:
+        upload_step = f"  [{_C_SAND}]phi upload {out}/[/]"
+
+    console.print(f"""
+[bold]── Tutorial: PD-L1 binder scoring pipeline ──────────────────[/]
+
+You have {len(files)} example binder structures in [{_C_BLUE}]{out}/[/].
+
+[bold]Step 1 — Upload[/]
+{upload_step}
+
+[bold]Step 2 — Run the filter pipeline[/]
+  [{_C_SAND}]phi filter --preset default --wait[/]
+
+  Runs:  ProteinMPNN → ESMFold → AlphaFold2 → score
+  Typical runtime: 10–30 min for {len(files)} structures.
+
+[bold]Step 3 — View scores[/]
+  [{_C_SAND}]phi scores[/]
+
+[bold]Step 4 — Download results[/]
+  [{_C_SAND}]phi download --out ./results[/]
+
+[bold]Dashboard[/]
+  [{_C_BLUE}]https://design.dynotx.com/dashboard[/]
+""")
diff --git a/src/phi/display.py b/src/phi/display.py
index bc0b2bd..424410e 100644
--- a/src/phi/display.py
+++ b/src/phi/display.py
@@ -272,7 +272,8 @@ def _print_status(s: dict) -> None:
         table.add_column("filename", style=_C_BLUE, no_wrap=True)
         table.add_column("type", style="dim")
         for f in files[:10]:
-            fname = f.get("filename") or f.get("gcs_url", "?")
+            raw = f.get("filename") or f.get("gcs_url", "?")
+            fname = raw.split("/")[-1] if raw.startswith("gs://") else raw
             ftype = f.get("artifact_type", "")
             table.add_row(fname, ftype)
         if len(files) > 10:
diff --git a/src/phi/parser.py b/src/phi/parser.py
index c0b04ba..2111784 100644
--- a/src/phi/parser.py
+++ b/src/phi/parser.py
@@ -4,15 +4,16 @@
 from phi.config import _FILTER_PRESETS, POLL_INTERVAL
 
 _CLI_EPILOG = """\
+Quick start:
+  phi tutorial               # download example structures + print step-by-step guide
+  phi filter --preset default --wait
+  phi scores
+  phi download --out ./results
+
 Fetch and prepare target structures:
   phi fetch --pdb 4ZQK --chain A --residues 56-290 --out target.pdb
   phi fetch --uniprot Q9NZQ7 --trim-low-confidence 70 --upload
 
-Design (backbone generation):
-  phi design      --target-pdb target.pdb --hotspots A45,A67 --num-designs 50
-  phi design      --length 80 --num-designs 20
-  phi boltzgen    --yaml design.yaml --protocol protein-anything --num-designs 10
-
 Validation (fold + score):
   phi esmfold     --fasta sequences.fasta
   phi alphafold   --fasta complex.fasta
@@ -20,10 +21,10 @@
   phi esm2        --fasta sequences.fasta
   phi boltz       --fasta complex.fasta
 
-Batch filter pipeline (100-50,000 designs):
-  phi upload --dir ./designs/ --file-type pdb
-  phi filter --dataset-id <id> --preset default --wait
-  phi download   --out ./results
+Batch filter pipeline:
+  phi upload ./designs/
+  phi filter --preset default --wait
+  phi download --out ./results
 
 Dataset management:
   phi datasets               # list your datasets
@@ -253,6 +254,17 @@ def build_parser() -> argparse.ArgumentParser:
     )
     sub = root.add_subparsers(dest="command", required=True)
 
+    p = sub.add_parser(
+        "tutorial",
+        help="Download example structures and print a step-by-step scoring walkthrough",
+    )
+    p.add_argument(
+        "--out",
+        metavar="DIR",
+        default="examples",
+        help="Directory to download example files into (default: ./examples)",
+    )
+
     p = sub.add_parser("login", help="Verify API key and print connection + identity details")
     p.add_argument("--json", action="store_true")