Add test inventory + checkpoint test runner

docs/testing/test_inventory.md: honest, current list of what automated
coverage actually exists (38 backend pytest tests across 2 files, 0 Godot
tests merged yet) and what doesn't (no integration/endpoint tests at all,
no coverage yet for the in-review RS256/moderation work) -- verified
against a genuinely fresh checkout + fresh Postgres, not just read off the
file tree.

scripts/run_checkpoint_tests.py: runs the whole inventory in one shot --
throwaway Postgres, migrations, pytest, then starts the backend server and
runs every scripts/smoke_test_*.gd against it headlessly. Verified working
both with and without a smoke test present (temporarily tested against
PC1's not-yet-merged scripts/api_client.gd + smoke_test_api.gd, removed
before this commit -- that PR carries its own files).

docs/testing/README.md: what running this at a Package Checkpoint review
does and doesn't prove, and how to grow the inventory as real coverage
gets added.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

.gitignore

@@ -8,6 +8,3 @@ export/
 # OS
 .DS_Store
 Thumbs.db
-
-# Godot UID files (regenerated from path)
-*.uid

docs/testing/README.md

@@ -0,0 +1,50 @@
+# Testing at Package Checkpoints
+
+A Package Checkpoint (PC1, PC2, ...) is a periodic review point — every few
+real milestones, the client and server builds get packaged together and
+independently reviewed before anyone calls anything actually done. See
+`project_agp` in Claude's memory, or ask, for the full definition; this
+doc is just the mechanical "how do I run the tests" part of that review.
+
+## Running the suite
+
+```
+python scripts/run_checkpoint_tests.py
+```
+
+Brings up a throwaway Postgres, runs Alembic migrations against it, runs the
+backend pytest suite, starts the backend server, and runs every Godot
+headless smoke test (`scripts/smoke_test_*.gd`) against it. Tears the
+Postgres container down afterward (`--keep-postgres` to skip that if you
+want to poke at the DB after a failure).
+
+If port 5432 is already taken on the machine, pass `--postgres-port` with
+something else.
+
+Exit code is 0 only if everything that actually ran passed — suites with
+nothing to run yet (no Godot binary found, no smoke test scripts yet) are
+reported SKIPPED, not treated as a failure.
+
+## What this does and doesn't prove
+
+Passing this script means: migrations apply cleanly to a genuinely empty
+database, the backend's unit-level logic is correct, and whatever Godot
+smoke tests exist actually round-trip against a live backend. It does
+**not** mean full endpoint/integration coverage exists — see
+`test_inventory.md` for the current, honest list of what's tested and what
+isn't. A checkpoint review should still involve reading the actual diffs
+for anything this script's current coverage doesn't reach, not just running
+green and calling it done.
+
+## Growing this over time
+
+As real integration tests get added (endpoint tests hitting the FastAPI app,
+more Godot smoke coverage beyond `health_check`), they should land under
+`backend/tests/` / `scripts/smoke_test_*.gd` respectively — this script
+picks up new files automatically, nothing to register by hand. Update
+`test_inventory.md`'s tables when you do, so the inventory stays a source
+of truth instead of quietly going stale.
+
+Eventual goal (not implemented yet): gate merges to `main` on this script
+passing, the same way a real CI pipeline would. For now it's a manual step
+run at checkpoint review time.

docs/testing/test_inventory.md

@@ -0,0 +1,78 @@
+# Test Inventory
+
+What automated coverage actually exists, what it actually verifies, and what
+doesn't have coverage yet. Keep this current as tests are added — it's the
+thing a Package Checkpoint review (see `docs/testing/README.md`) reads first
+to know what "run the tests" actually checked.
+
+Run everything below in one shot with `scripts/run_checkpoint_tests.py`.
+
+Last verified against `main`: 2026-08-19 (commit `1c75367`).
+
+## Backend — `backend/tests/` (pytest)
+
+Run: `cd backend && python -m pytest -v` (needs `DATABASE_URL` pointed at a
+reachable Postgres — see the runner script; most of these tests don't
+actually touch the DB, but the app's import chain requires a valid URL to
+construct).
+
+**Current total: 38 tests, all passing on a clean `main` checkout.**
+
+| File | Tests | What it covers |
+|---|---|---|
+| `test_auth_age_gate.py` | 10 | `calculate_age()` date-math correctness (exact boundary, leap years, birthday-not-yet-this-year) + `RegisterRequest` date-of-birth validation (parametrized valid/invalid date rejection). Backs the 18+ registration gate (PR #23). |
+| `test_gltf_metadata.py` | 21 | glTF/GLB binary + JSON parsing: magic-byte detection, chunk-padding handling, mesh/animation/texture-reference metadata extraction, malformed-input rejection. Backs the asset-import pipeline (PR #27). |
+
+**What these are, honestly:** pure unit tests of standalone functions
+(`calculate_age`, the glTF parser, Pydantic validators). None of them spin
+up the FastAPI app or hit a real endpoint — they run in ~1s and don't
+actually need Postgres reachable at all (the import chain just needs a
+syntactically valid `DATABASE_URL`, nothing connects during collection).
+
+**Known coverage gaps** (real gaps, not placeholders — nothing enforces filling
+these yet):
+- No integration/endpoint tests at all — `/register`, `/login`, `/refresh`,
+  `/worlds`, `/assets/*` have zero tests that actually call them through
+  the FastAPI app (e.g. via `httpx.AsyncClient` + `ASGITransport`) against a
+  real (or test) database. Everything currently shipped works because it
+  was manually verified at merge time (checkpoint review), not because a
+  test asserts it.
+- No tests for the RS256 auth upgrade (key generation, refresh-token
+  issuance/persistence) — PR #30, pending review as of this writing,
+  manually verified (correct key usage, all 38 existing tests still pass)
+  but not covered by an automated test of its own yet.
+- No tests for the moderation schema/migration (in progress as of this
+  writing) beyond "the migration runs."
+
+## Godot client — `scripts/smoke_test_*.gd`
+
+**Nothing merged into `main` yet as of this writing** — `sandbox_probe/test_script.gd`
+is the only Godot script on `main` right now, and it's **not a real
+test** (a leftover manual probe, returns a hardcoded string, asserts
+nothing). Don't count it toward coverage; delete or repurpose it if anyone
+actually needs a sandbox probe again.
+
+`scripts/smoke_test_api.gd` (PC1, PR #29, pending review as of this writing)
+will be the first real one once merged — `ApiClient` autoload's
+`health_check()` against a **live backend** (set `AGP_API_BASE_URL` to point
+at one), verified by hand end-to-end before opening the PR: exercises the
+real HTTP request path (`HTTPRequest` node lifecycle, JSON parse, the
+result-dictionary contract), not just that the script compiles. Exits 1 on
+failure, so it's CI/script-friendly.
+
+Run (once merged): `godot --headless --path . --import` once (builds the
+class cache — skip this and every `class_name`-declared class falsely
+errors as undeclared), then for each `scripts/smoke_test_*.gd`:
+`godot --headless --path . --script res://scripts/<name>.gd`
+
+**Known gap once it lands:** only `health_check()` is exercised end-to-end.
+`login`, `register`, `get_worlds`, `get_me` are implemented (see
+`scripts/api_client.gd`) but have no smoke coverage yet — each would need a
+real user/world fixture in the target backend to test against meaningfully.
+
+## What "full coverage" should eventually mean here
+
+Not defined precisely yet — the honest target as of 2026-08-19 is just
+"more than pure-function unit tests." As real integration/endpoint tests get
+added, update the tables above rather than letting this doc drift stale;
+a wrong inventory is worse than no inventory.

export_presets.cfg

@@ -1,74 +0,0 @@
-[preset.0]
-
-name="Linux"
-platform="Linux/X11"
-runnable=true
-advanced_options=false
-dedicated_server=false
-custom_features=""
-export_filter="all_resources"
-include_filter=""
-exclude_filter=""
-export_path="export/linux/agp-game.x86_64"
-encryption_include_filters=""
-encryption_exclude_filters=""
-encrypt_pck=false
-encrypt_directory=false
-script_export_mode=2
-
-[preset.0.options]
-
-custom_template/debug=""
-custom_template/release=""
-debug/export_console_wrapper=1
-binary_format/embed_pck=false
-texture_format/bptc=true
-texture_format/s3tc=true
-texture_format/etc=false
-texture_format/etc2=false
-binary_format/architecture="x86_64"
-ssh_remote_deploy/enabled=false
-
-[preset.1]
-
-name="Windows Desktop"
-platform="Windows Desktop"
-runnable=true
-advanced_options=false
-dedicated_server=false
-custom_features=""
-export_filter="all_resources"
-include_filter=""
-exclude_filter=""
-export_path="export/windows/agp-game.exe"
-encryption_include_filters=""
-encryption_exclude_filters=""
-encrypt_pck=false
-encrypt_directory=false
-script_export_mode=2
-
-[preset.1.options]
-
-custom_template/debug=""
-custom_template/release=""
-debug/export_console_wrapper=1
-binary_format/embed_pck=false
-texture_format/bptc=true
-texture_format/s3tc=true
-texture_format/etc=false
-texture_format/etc2=false
-binary_format/architecture="x86_64"
-codesign/enable=false
-application/icon=""
-application/console_wrapper_icon=""
-application/icon_interpolation=4
-application/file_version=""
-application/product_version=""
-application/company_name=""
-application/product_name=""
-application/file_description=""
-application/copyright=""
-application/trademarks=""
-application/export_angle=0
-application/export_d3d12=0
-d3d12/agility_sdk_multizip=0

project.godot

@@ -9,12 +9,9 @@ config_version=5
 config/name="AGP-Game"
 run/main_scene="res://scenes/main_menu.tscn"
 config/features=PackedStringArray("4.6.2")
-agp/api_base_url="http://127.0.0.1:8000"
 
 [autoload]
 
-ApiClient="*res://scripts/api_client.gd"
-
 [display]
 
 window/size/viewport_width=1920

scripts/api_client.gd

@@ -1,292 +0,0 @@
-extends Node
-## Central HTTP client singleton (autoload name: "ApiClient").
-##
-## Every network call in the game flows through this class. Callers use the
-## high-level helpers ([method login], [method register], [method get_worlds],
-## [method health_check], [method get_me]) or the generic [method request]
-## primitive. All results conform to the result-dictionary contract below.
-##
-## [b]Result Dictionary Contract[/b]
-## Every public method returns (after `await`) a Dictionary with this shape:
-##
-##     {
-##         "ok":     bool,     # true for HTTP 2xx, false otherwise
-##         "status": int,      # HTTP status code (0 on network failure)
-##         "data":   Variant,  # parsed JSON body on success (null otherwise)
-##         "detail": String,   # human-readable message
-##     }
-##
-## Network failures (timeout / DNS / refused connection) return `status == 0`
-## with a `detail` describing the transport failure. Non-2xx responses return
-## the real HTTP status and the backend's `detail` field from the JSON error
-## body when present. No caller can crash on a failed request.
-##
-## [b]Token storage[/b]
-## JWTs are kept in memory. [method set_token] is the single seam for future
-## disk persistence (override it to save to `user://`).
-##
-## [b]Usage[/b]
-##     var result: Dictionary = await ApiClient.login("alice", "secret")
-##     if result.ok:
-##         print("Logged in!")
-##     else:
-##         print("Error: ", result.detail)
-
-# ---------------------------------------------------------------------------
-# Token storage
-# ---------------------------------------------------------------------------
-
-var _access_token: String = ""
-var _token_type: String = "Bearer"
-
-## Full Authorization header value, e.g. "Bearer eyJ...". Empty if no token.
-var bearer_token: String:
-	get:
-		return _token_type + " " + _access_token if _access_token != "" else ""
-
-
-## Whether an access token is currently stored.
-func has_token() -> bool:
-	return _access_token != ""
-
-
-## Set the access token (called after login / register / token refresh).
-## This is the persistence seam: override to also write to disk.
-func set_token(access_token: String, token_type: String = "Bearer") -> void:
-	_access_token = access_token
-	_token_type = token_type
-
-
-## Clear stored tokens.
-func clear_token() -> void:
-	_access_token = ""
-	_token_type = "Bearer"
-
-
-# ---------------------------------------------------------------------------
-# Configuration
-# ---------------------------------------------------------------------------
-
-var _base_url: String = "http://127.0.0.1:8000"
-
-## Request timeout in seconds. 0 disables the timeout.
-var timeout_seconds: float = 10.0
-
-
-func _ready() -> void:
-	# Environment variable override (e.g. for headless smoke tests against a
-	# backend on a non-default port). Falls back to the ProjectSetting, then
-	# to the documented default.
-	var env_url := OS.get_environment("AGP_API_BASE_URL")
-	if env_url != "":
-		_base_url = env_url.rstrip("/")
-	else:
-		_base_url = str(ProjectSettings.get_setting(
-			"application/agp/api_base_url",
-			"http://127.0.0.1:8000"
-		)).rstrip("/")
-
-
-# ---------------------------------------------------------------------------
-# Core request primitive
-# ---------------------------------------------------------------------------
-
-## Generic request primitive. Awaitable via `await`.
-##
-## Uses a dedicated HTTPRequest node per call so concurrent requests can never
-## cross-talk, and so the result maps 1:1 to the awaiting caller.
-##
-## @param method       HTTPClient.Method enum value (GET/POST/PUT/PATCH/DELETE).
-## @param path         URL path, e.g. "/api/health".
-## @param body         Request body string (JSON or form-encoded).
-## @param content_type "application/json" or "application/x-www-form-urlencoded".
-## @param require_auth If true, attach the stored bearer token when present.
-## @return Result dictionary conforming to the contract.
-func request(
-	method: HTTPClient.Method,
-	path: String,
-	body: String = "",
-	content_type: String = "application/json",
-	require_auth: bool = false
-) -> Dictionary:
-	var http := HTTPRequest.new()
-	http.timeout = timeout_seconds
-	http.max_redirects = 8
-	http.accept_gzip = true
-	add_child(http)
-
-	var headers := PackedStringArray()
-	headers.append("Accept: application/json")
-	if content_type != "":
-		headers.append("Content-Type: " + content_type)
-	if require_auth and has_token():
-		headers.append("Authorization: " + bearer_token)
-
-	var url := _base_url + path
-	var err := http.request(url, headers, method, body)
-	if err != OK:
-		http.queue_free()
-		return _make_result(false, 0, null, "Failed to start request: " + _error_string(err))
-
-	# request_completed emits (result, response_code, headers, body).
-	var args: Array = await http.request_completed
-	var result_code: int = args[0]
-	var response_code: int = args[1]
-	var response_body: PackedByteArray = args[3]
-	http.queue_free()
-
-	# Transport-level failure (timeout / DNS / refused / TLS, etc.).
-	if result_code != HTTPRequest.RESULT_SUCCESS:
-		return _make_result(false, 0, null, _transport_error_detail(result_code))
-
-	# Server accepted the connection but returned no HTTP status line.
-	if response_code == 0:
-		return _make_result(false, 0, null, "No response received from server.")
-
-	var json_string := response_body.get_string_from_utf8()
-	var ok := response_code >= 200 and response_code < 300
-
-	if ok:
-		var data: Variant = _parse_json(json_string)
-		return _make_result(true, response_code, data, "OK")
-
-	# Non-2xx: surface the backend's "detail" message when available.
-	var detail := "HTTP error %d" % response_code
-	var parsed: Variant = _parse_json(json_string)
-	if parsed is Dictionary and (parsed as Dictionary).has("detail"):
-		detail = str((parsed as Dictionary)["detail"])
-	return _make_result(false, response_code, null, detail)
-
-
-# ---------------------------------------------------------------------------
-# High-level helpers
-# ---------------------------------------------------------------------------
-
-## Log in with username/password (OAuth2PasswordRequestForm).
-## On success stores the returned access token.
-func login(username: String, password: String) -> Dictionary:
-	var form := "username=" + username.uri_encode() + "&password=" + password.uri_encode()
-	var result := await request(
-		HTTPClient.METHOD_POST,
-		"/api/auth/login",
-		form,
-		"application/x-www-form-urlencoded",
-		false
-	)
-	_store_token_from_result(result)
-	return result
-
-
-## Register a new user. On success stores the returned access token.
-## @param date_of_birth ISO date string "YYYY-MM-DD".
-func register(
-	username: String,
-	email: String,
-	password: String,
-	date_of_birth: String
-) -> Dictionary:
-	var payload := {
-		"username": username,
-		"email": email,
-		"password": password,
-		"date_of_birth": date_of_birth,
-	}
-	var result := await request(
-		HTTPClient.METHOD_POST,
-		"/api/auth/register",
-		JSON.stringify(payload),
-		"application/json",
-		false
-	)
-	_store_token_from_result(result)
-	return result
-
-
-## List worlds visible to the current user (requires auth).
-## @param visibility Optional filter: "public", "private", "friends_only".
-func get_worlds(skip: int = 0, limit: int = 50, visibility: String = "") -> Dictionary:
-	var path := "/api/worlds/?skip=%d&limit=%d" % [skip, limit]
-	if visibility != "":
-		path += "&visibility=" + visibility.uri_encode()
-	return await request(HTTPClient.METHOD_GET, path, "", "application/json", true)
-
-
-## Ping the backend health endpoint (no auth).
-func health_check() -> Dictionary:
-	return await request(HTTPClient.METHOD_GET, "/api/health", "", "application/json", false)
-
-
-## Fetch the current user's profile from the stored JWT (requires auth).
-func get_me() -> Dictionary:
-	return await request(HTTPClient.METHOD_GET, "/api/users/me", "", "application/json", true)
-
-
-# ---------------------------------------------------------------------------
-# Internals
-# ---------------------------------------------------------------------------
-
-func _store_token_from_result(result: Dictionary) -> void:
-	if result.ok and result.data is Dictionary:
-		var token: String = str((result.data as Dictionary).get("access_token", ""))
-		if token != "":
-			set_token(token)
-
-
-func _parse_json(text: String) -> Variant:
-	if text == "":
-		return null
-	var parsed: Variant = JSON.parse_string(text)
-	if parsed == null:
-		# null is ambiguous (valid JSON "null" vs parse failure). Fall back to
-		# the raw string so callers never lose the body entirely.
-		return text if text != "null" else null
-	return parsed
-
-
-func _make_result(ok: bool, status: int, data: Variant, detail: String) -> Dictionary:
-	return {
-		"ok": ok,
-		"status": status,
-		"data": data,
-		"detail": detail,
-	}
-
-
-func _transport_error_detail(result_code: int) -> String:
-	match result_code:
-		HTTPRequest.RESULT_CANT_CONNECT:
-			return "Could not connect to server. Is it running?"
-		HTTPRequest.RESULT_CANT_RESOLVE:
-			return "Could not resolve server hostname."
-		HTTPRequest.RESULT_CONNECTION_ERROR:
-			return "Connection error."
-		HTTPRequest.RESULT_TLS_HANDSHAKE_ERROR:
-			return "TLS handshake failed."
-		HTTPRequest.RESULT_NO_RESPONSE:
-			return "No response from server."
-		HTTPRequest.RESULT_BODY_SIZE_LIMIT_EXCEEDED:
-			return "Response body too large."
-		HTTPRequest.RESULT_BODY_DECOMPRESS_FAILED:
-			return "Failed to decompress response body."
-		HTTPRequest.RESULT_REQUEST_FAILED:
-			return "Request failed."
-		HTTPRequest.RESULT_TIMEOUT:
-			return "Request timed out."
-		_:
-			return "Network error (code %d)." % result_code
-
-
-func _error_string(err_code: int) -> String:
-	match err_code:
-		OK:
-			return "OK"
-		ERR_UNCONFIGURED:
-			return "HTTPRequest not configured (not in tree)."
-		ERR_BUSY:
-			return "HTTPRequest is busy with a previous request."
-		ERR_INVALID_PARAMETER:
-			return "Invalid URL format."
-		ERR_CANT_CONNECT:
-			return "Cannot connect to host."
-		_:
-			return "Unknown error (%d)." % err_code

scripts/build.sh

@@ -1,42 +0,0 @@
-#!/usr/bin/env bash
-# Headless CI export for Linux and Windows.
-#
-# Requires:
-#   - Godot 4.6.2 (the exact version this project + export_presets.cfg were
-#     built against; a different Godot minor version's export templates are
-#     not guaranteed compatible)
-#   - The matching export templates installed at
-#     ~/.local/share/godot/export_templates/4.6.2.stable/
-#     (download: https://godotengine.org/download/archive/4.6.2-stable/,
-#     "Export templates" -- NOT the same as the editor download)
-#   - godot on PATH (or set GODOT_BIN below)
-#
-# Usage: scripts/build.sh
-# Output: export/linux/agp-game.x86_64, export/windows/agp-game.exe
-
-set -euo pipefail
-cd "$(dirname "$0")/.."
-
-GODOT_BIN="${GODOT_BIN:-godot}"
-if ! command -v "$GODOT_BIN" >/dev/null 2>&1; then
-    if [ -x "$HOME/godot/godot" ]; then
-        GODOT_BIN="$HOME/godot/godot"
-    else
-        echo "godot binary not found (checked PATH and ~/godot/godot)." >&2
-        exit 1
-    fi
-fi
-
-echo "Godot version: $("$GODOT_BIN" --version)"
-
-mkdir -p export/linux export/windows
-
-echo "=== Exporting Linux/X11 ==="
-"$GODOT_BIN" --headless --path . --export-release "Linux" export/linux/agp-game.x86_64
-
-echo "=== Exporting Windows Desktop ==="
-"$GODOT_BIN" --headless --path . --export-release "Windows Desktop" export/windows/agp-game.exe
-
-echo ""
-echo "Build artifacts:"
-ls -la export/linux export/windows

scripts/run_checkpoint_tests.py

@@ -0,0 +1,170 @@
+#!/usr/bin/env python3
+"""Run the full automated test inventory for a Package Checkpoint review.
+
+See docs/testing/test_inventory.md for what each suite actually covers (and
+doesn't) -- keep that doc in sync as tests are added, this script just runs
+whatever's there.
+
+Usage:
+    python scripts/run_checkpoint_tests.py [--keep-postgres] [--postgres-port PORT]
+
+Exits 0 if everything that ran passed, 1 otherwise. Suites with nothing to
+run (e.g. no scripts/smoke_test_*.gd yet, or no Godot binary found) are
+reported as SKIPPED, not counted as a failure.
+"""
+import argparse
+import glob
+import os
+import shutil
+import subprocess
+import sys
+import time
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parent.parent
+BACKEND = ROOT / "backend"
+
+
+def run(cmd, cwd=None, env=None, capture=False):
+    print(f"$ {' '.join(cmd)}" + (f"  (cwd={cwd})" if cwd else ""))
+    return subprocess.run(cmd, cwd=cwd, env=env, capture_output=capture, text=True)
+
+
+def find_godot():
+    for candidate in ("godot", str(Path.home() / "godot" / "godot")):
+        if shutil.which(candidate) or Path(candidate).is_file():
+            return candidate
+    return None
+
+
+def wait_for_http(url: str, timeout_s: int = 20) -> bool:
+    import urllib.request
+    import urllib.error
+
+    deadline = time.time() + timeout_s
+    while time.time() < deadline:
+        try:
+            with urllib.request.urlopen(url, timeout=2) as resp:
+                if resp.status == 200:
+                    return True
+        except (urllib.error.URLError, ConnectionError, OSError):
+            pass
+        time.sleep(1)
+    return False
+
+
+def main():
+    parser = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
+    parser.add_argument("--keep-postgres", action="store_true",
+                         help="Don't tear down the checkpoint Postgres container afterward")
+    parser.add_argument("--postgres-port", default=os.environ.get("CHECKPOINT_POSTGRES_PORT", "5432"),
+                         help="Host port for the checkpoint's own Postgres container (default 5432 -- "
+                              "override if that's already in use on this machine)")
+    parser.add_argument("--backend-port", default=os.environ.get("CHECKPOINT_BACKEND_PORT", "18080"),
+                         help="Port to run the backend server on for the Godot smoke tests")
+    args = parser.parse_args()
+
+    results = {}
+
+    # --- Backend: bring up Postgres ---
+    compose_env = os.environ.copy()
+    compose_env["COMPOSE_PROJECT_NAME"] = "agp-checkpoint"
+    compose_env["POSTGRES_HOST_PORT"] = args.postgres_port
+    compose_env["POSTGRES_PORT"] = args.postgres_port  # some branches use this name instead, see tools/definitions.py
+    print("=== Backend: bringing up Postgres ===")
+    run(["docker", "compose", "up", "-d", "postgres"], cwd=BACKEND, env=compose_env)
+
+    print("Waiting for Postgres to be ready...")
+    pg_ready = False
+    for _ in range(20):
+        r = run(["docker", "compose", "exec", "-T", "postgres", "pg_isready", "-U", "agp_user", "-d", "agp_db"],
+                cwd=BACKEND, env=compose_env, capture=True)
+        if r.returncode == 0:
+            pg_ready = True
+            break
+        time.sleep(1)
+    if not pg_ready:
+        print("Postgres never became ready -- aborting.")
+        sys.exit(1)
+
+    db_url = f"postgresql+asyncpg://agp_user:agp_password@localhost:{args.postgres_port}/agp_db"
+    backend_env = os.environ.copy()
+    backend_env["DATABASE_URL"] = db_url
+
+    print("\n=== Backend: alembic upgrade head ===")
+    r = run([sys.executable, "-m", "alembic", "upgrade", "head"], cwd=BACKEND, env=backend_env)
+    results["migrations (fresh DB)"] = r.returncode == 0
+
+    print("\n=== Backend: pytest ===")
+    r = run([sys.executable, "-m", "pytest", "-v"], cwd=BACKEND, env=backend_env)
+    results["backend pytest"] = r.returncode == 0
+
+    # --- Godot: headless smoke tests (need a live backend server) ---
+    godot_bin = find_godot()
+    smoke_tests = sorted(glob.glob(str(ROOT / "scripts" / "smoke_test_*.gd")))
+
+    if not godot_bin:
+        print("\n=== Godot: binary not found (checked PATH and ~/godot/godot), skipping smoke tests ===")
+        results["godot smoke tests"] = None
+    elif not smoke_tests:
+        print("\n=== Godot: no scripts/smoke_test_*.gd found yet, nothing to run ===")
+        results["godot smoke tests"] = None
+    else:
+        print(f"\n=== Backend: starting server on :{args.backend_port} for Godot smoke tests ===")
+        server_proc = subprocess.Popen(
+            [sys.executable, "-m", "uvicorn", "app.main:app", "--host", "127.0.0.1", "--port", args.backend_port],
+            cwd=BACKEND, env=backend_env,
+            stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL,
+        )
+        try:
+            base_url = f"http://127.0.0.1:{args.backend_port}"
+            if not wait_for_http(f"{base_url}/api/health"):
+                print("Backend server never became healthy -- skipping Godot smoke tests.")
+                results["godot smoke tests"] = False
+            else:
+                godot_env = os.environ.copy()
+                godot_env["AGP_API_BASE_URL"] = base_url
+
+                print("=== Godot: import pass (builds the class cache) ===")
+                run([godot_bin, "--headless", "--path", str(ROOT), "--import"], env=godot_env)
+                run([godot_bin, "--headless", "--path", str(ROOT), "--quit"], env=godot_env)
+
+                godot_ok = True
+                for test in smoke_tests:
+                    rel = os.path.relpath(test, ROOT)
+                    print(f"\n--- {rel} ---")
+                    r = run([godot_bin, "--headless", "--path", str(ROOT), "--script", f"res://{rel}"], env=godot_env)
+                    if r.returncode != 0:
+                        godot_ok = False
+                results["godot smoke tests"] = godot_ok
+        finally:
+            server_proc.terminate()
+            try:
+                server_proc.wait(timeout=10)
+            except subprocess.TimeoutExpired:
+                server_proc.kill()
+
+    if not args.keep_postgres:
+        print("\n=== Backend: tearing down checkpoint Postgres ===")
+        run(["docker", "compose", "down", "-v", "postgres"], cwd=BACKEND, env=compose_env)
+
+    # --- Summary ---
+    print("\n" + "=" * 50)
+    print("SUMMARY")
+    print("=" * 50)
+    overall_ok = True
+    for name, ok in results.items():
+        if ok is None:
+            print(f"  {name}: SKIPPED")
+        elif ok:
+            print(f"  {name}: PASS")
+        else:
+            print(f"  {name}: FAIL")
+            overall_ok = False
+
+    print("\nSee docs/testing/test_inventory.md for what each suite actually covers (and doesn't).")
+    sys.exit(0 if overall_ok else 1)
+
+
+if __name__ == "__main__":
+    main()

scripts/smoke_test_api.gd

@@ -1,46 +0,0 @@
-extends SceneTree
-## Headless smoke test for the ApiClient autoload.
-##
-## Run with:
-##     godot --headless --path . --script res://scripts/smoke_test_api.gd
-##
-## Requires the backend to be running (default base URL
-## `application/agp/api_base_url` -> http://127.0.0.1:8000).
-##
-## Calls the real autoload's `health_check()` (which exercises the generic
-## `request()` primitive) and exits 0 on a `status == "ok"` response, 1
-## otherwise. This script is intentionally NOT wired into any game scene.
-
-
-func _initialize() -> void:
-	_run()
-
-
-func _run() -> void:
-	# _initialize() fires before the SceneTree has processed its first frame,
-	# so a freshly add_child()'d HTTPRequest node isn't reliably considered
-	# "inside tree" yet (request_raw fails with !is_inside_tree()). One frame
-	# of delay is enough; this only matters for --script mode, not a real
-	# running game.
-	await process_frame
-
-	# Godot's --script/SceneTree runner doesn't resolve autoload singletons
-	# as bare global identifiers at compile time (the node exists at
-	# runtime -- root.has_node("ApiClient") is true -- but referencing it as
-	# a plain identifier fails with "Identifier not found: ApiClient" during
-	# script compilation). get_node sidesteps compile-time resolution
-	# entirely. This only affects --script mode; a real game scene resolves
-	# the bare autoload name normally.
-	var api_client := root.get_node("ApiClient")
-	print("SMOKE: calling ApiClient.health_check() ...")
-	var result: Dictionary = await api_client.health_check()
-	print("SMOKE: result = ", result)
-
-	if result.ok \
-			and result.data is Dictionary \
-			and str((result.data as Dictionary).get("status", "")) == "ok":
-		print("SMOKE RESULT: PASS (status=ok)")
-		quit(0)
-	else:
-		printerr("SMOKE RESULT: FAIL")
-		quit(1)