feat(#224): multi-region live pricing sync + sync-pricing CLI; seed is test-only

DESIGN_PRINCIPLES.md

@@ -211,6 +211,6 @@ If the TypeScript surface later needs a standalone engine, it reimplements the s
 
 Cloud prices change monthly. Hard-coded prices in source code go stale. The pricing layer exposes a query interface and the engine never knows or cares where the data comes from.
 
-The catalog must handle tiered pricing (e.g., Lambda GB-seconds has three price tiers), free tiers (first 1M requests free), and regional variation. It must work offline after an initial seed. Prices refresh on a schedule — cloud pricing changes monthly at most.
+The catalog must handle tiered pricing (e.g., Lambda GB-seconds has three price tiers), free tiers (first 1M requests free), and regional variation across all services and regions. Prices are fetched live from the pricing source and refresh on a schedule — cloud pricing changes monthly at most. The bundled seed price list is a test fixture and offline fallback, not a normal setup step.
 
 A provider plugin architecture is premature. Use a normalized multi-cloud source until it doesn't cover a needed provider.

README.md

@@ -39,6 +39,18 @@ infra-cost-model what-if model-a.yaml --compare model-b.yaml \
 infra-cost-model graph model.yaml
 ```
 
+## Pricing data
+
+Prices are fetched live from the [Infracost Cloud Pricing API](https://www.infracost.io/docs/), covering all supported services and all regions:
+
+```bash
+# Authenticate once: set INFRACOST_API_KEY, or run `infracost auth login`
+infra-cost-model sync-pricing                              # all services, all regions
+infra-cost-model sync-pricing --region us-east-1 --region eu-west-1
+```
+
+The bundled `data/seed/aws_pricelist_seed.json` is a small us-east-1 fixture used by the test suite only — it is **not** a setup step for users, and `seed-pricing` exists purely for offline/testing.
+
 ## References
 
 - Leitner, Cito & Stöckli. "Modelling and Managing Deployment Costs of Microservice-Based Cloud Applications." *UCC 2016*. DOI: 10.1145/2996890.2996901

infra_cost_model/cli.py

@@ -74,8 +74,22 @@ def _build_parser() -> argparse.ArgumentParser:
     p_extract.add_argument("--json", action="store_true", help="Output in JSON format")
     p_extract.set_defaults(func=cmd_extract)
 
-    # seed-pricing
-    p_seed = sub.add_parser("seed-pricing", help="Seed pricing cache from seed file")
+    # sync-pricing (live, all services/regions — the normal way to populate prices)
+    p_sync = sub.add_parser(
+        "sync-pricing",
+        help="Fetch live prices from Infracost for all services and regions")
+    p_sync.add_argument("services", nargs="*", metavar="<metric>",
+                        help="Specific catalog metrics to sync (default: all)")
+    p_sync.add_argument("--region", action="append", dest="regions", metavar="REGION",
+                        help="Region to sync (repeatable). Default: all known regions")
+    p_sync.add_argument("--vendor", default="aws", metavar="VENDOR",
+                        help="Cloud vendor (default: aws)")
+    p_sync.set_defaults(func=cmd_sync_pricing)
+
+    # seed-pricing (testing/offline only — NOT a normal user step; see sync-pricing)
+    p_seed = sub.add_parser(
+        "seed-pricing",
+        help="(testing/offline) Load the bundled seed price fixtures into the cache")
     p_seed.add_argument("services", nargs="*", metavar="<service>",
                         help="Specific services to seed (default: all)")
     p_seed.add_argument("--all", action="store_true", default=True,
@@ -331,8 +345,42 @@ def cmd_analyze(args: argparse.Namespace) -> int:
         return 1
 
 
+def cmd_sync_pricing(args: argparse.Namespace) -> int:
+    """Fetch live prices from Infracost across all services and regions.
+
+    This is the normal way to populate the pricing catalog. Requires an Infracost
+    credential (INFRACOST_API_KEY or `infracost auth login`); without one it falls
+    back to the bundled seed fixtures and warns.
+    """
+    from infra_cost_model.pricing.sources.infracost import (
+        sync_pricing_catalog, _REGION_PREFIX,
+    )
+
+    services = args.services if args.services else None
+    regions = args.regions if args.regions else sorted(_REGION_PREFIX)
+
+    try:
+        count, source = sync_pricing_catalog(
+            vendor=args.vendor, services=services, regions=regions)
+        if source == "infracost":
+            print(f"✓ Synced {count} prices from {source} across {len(regions)} region(s)")
+        else:
+            # No live credential → sync_pricing_catalog fell back to the us-east-1
+            # seed fixtures; don't claim the full region fan-out happened.
+            print(f"✓ Loaded {count} prices from {source} "
+                  f"(us-east-1 fallback — set INFRACOST_API_KEY for live, all-region pricing)")
+        return 0
+    except RuntimeError as e:
+        _print_stderr(f"Error: {e}")
+        return 1
+
+
 def cmd_seed_pricing(args: argparse.Namespace) -> int:
-    """Seed pricing catalog from seed file."""
+    """Load bundled seed price fixtures (testing/offline only).
+
+    Not a normal user step — real pricing comes from `sync-pricing`. The seed list
+    is a small us-east-1 fixture used by the test suite.
+    """
     from infra_cost_model.pricing.sources.infracost import seed_pricing_catalog
 
     services = args.services if args.services else None

infra_cost_model/pricing/sources/infracost.py

@@ -566,12 +566,16 @@ def _live_auth_intended(client: "InfracostClient") -> bool:
 
 
 def sync_pricing_catalog(vendor: str = "aws", services: list[str] = None,
-                         fallback: bool = False) -> tuple[int, str]:
+                         fallback: bool = False,
+                         regions: list[str] = None) -> tuple[int, str]:
     """Sync pricing into the cache, live from Infracost when authenticated.
 
-    Falls back to the bundled seed price list when there is no credential, but
-    emits a ``UserWarning`` when a credential WAS present and the live sync failed
-    — so a broken live path is never silently mistaken for success.
+    Fetches every descriptor's prices for each requested region (defaults to
+    us-east-1 for backward compatibility; the CLI ``sync-pricing`` command passes
+    the full region set so pricing covers all services and all regions). Falls
+    back to the bundled seed price list when there is no credential, but emits a
+    ``UserWarning`` when a credential WAS present and the live sync failed — so a
+    broken live path is never silently mistaken for success.
     """
     from infra_cost_model.pricing.cache import PricingCache
 
@@ -586,16 +590,18 @@ def sync_pricing_catalog(vendor: str = "aws", services: list[str] = None,
         return _sync_fallback(vendor, services, cache)
 
     metrics = services if services else list(METRIC_DESCRIPTORS.keys())
-    region = "us-east-1"
+    if not regions:
+        regions = ["us-east-1"]
     total = 0
     failures: list[str] = []
-    for metric in metrics:
-        if metric not in METRIC_DESCRIPTORS:
-            continue
-        try:
-            total += client.sync_to_cache(cache, metric, region, vendor)
-        except (RuntimeError, requests.RequestException, KeyError) as exc:
-            failures.append(f"{metric}: {exc}")
+    for region in regions:
+        for metric in metrics:
+            if metric not in METRIC_DESCRIPTORS:
+                continue
+            try:
+                total += client.sync_to_cache(cache, metric, region, vendor)
+            except (RuntimeError, requests.RequestException, KeyError) as exc:
+                failures.append(f"{region}/{metric}: {exc}")
 
     if total == 0:
         warnings.warn(

tests/test_cli.py

@@ -1646,3 +1646,45 @@ def test_coverage_pulumi_format(self):
             os.unlink(temp_yaml)
             os.unlink(temp_tf)
 
+
+
+def test_cli_sync_pricing_defaults_to_all_regions(monkeypatch):
+    """`sync-pricing` with no --region syncs every known region."""
+    import infra_cost_model.pricing.sources.infracost as ic
+    captured = {}
+
+    def fake_sync(vendor="aws", services=None, regions=None):
+        captured["vendor"] = vendor
+        captured["services"] = services
+        captured["regions"] = regions
+        return (42, "infracost")
+
+    monkeypatch.setattr(ic, "sync_pricing_catalog", fake_sync)
+    rc = main(["sync-pricing"])
+    assert rc == 0
+    assert captured["services"] is None
+    assert set(captured["regions"]) == set(ic._REGION_PREFIX)
+
+
+def test_cli_sync_pricing_explicit_regions(monkeypatch):
+    import infra_cost_model.pricing.sources.infracost as ic
+    captured = {}
+    monkeypatch.setattr(ic, "sync_pricing_catalog",
+                        lambda vendor="aws", services=None, regions=None:
+                            captured.update(regions=regions) or (1, "infracost"))
+    rc = main(["sync-pricing", "--region", "eu-west-1", "--region", "us-west-2"])
+    assert rc == 0
+    assert captured["regions"] == ["eu-west-1", "us-west-2"]
+
+
+def test_cli_sync_pricing_fallback_message(monkeypatch, capsys):
+    """When no credential → seed fallback, the message must not claim all regions."""
+    import infra_cost_model.pricing.sources.infracost as ic
+    monkeypatch.setattr(ic, "sync_pricing_catalog",
+                        lambda vendor="aws", services=None, regions=None: (14, "seed-pricelist"))
+    rc = main(["sync-pricing"])
+    assert rc == 0
+    out = capsys.readouterr().out
+    assert "seed-pricelist" in out
+    assert "region(s)" not in out  # must not overstate the fan-out
+    assert "fallback" in out.lower()

tests/test_infracost_client.py

@@ -436,3 +436,36 @@ def test_ap_region_prefixes_match_aws_codes():
     assert _REGION_PREFIX["ap-south-1"] == "APS3"      # Mumbai
     assert _REGION_PREFIX["ap-southeast-3"] == "APS4"  # Jakarta
     assert _REGION_PREFIX["ap-south-2"] == "APS5"      # Hyderabad
+
+
+# --- Multi-region live sync ----------------------------------------------------
+
+def test_sync_pricing_catalog_covers_multiple_regions(monkeypatch):
+    """sync_pricing_catalog fetches every metric for EACH requested region."""
+    _set_creds(monkeypatch)
+    calls = []
+    monkeypatch.setattr(ic.InfracostClient, "is_authenticated", lambda self: True)
+    monkeypatch.setattr(
+        ic.InfracostClient, "sync_to_cache",
+        lambda self, cache, usage_metric, region, vendor="aws":
+            calls.append((usage_metric, region)) or 1,
+    )
+    total, source = ic.sync_pricing_catalog(
+        services=["KMS-Key-Month"], regions=["us-east-1", "eu-west-1", "ap-south-1"])
+    assert source == "infracost"
+    assert total == 3
+    assert {r for _, r in calls} == {"us-east-1", "eu-west-1", "ap-south-1"}
+
+
+def test_sync_pricing_catalog_defaults_to_us_east_1(monkeypatch):
+    """No regions arg → us-east-1 only (backward compatible)."""
+    _set_creds(monkeypatch)
+    calls = []
+    monkeypatch.setattr(ic.InfracostClient, "is_authenticated", lambda self: True)
+    monkeypatch.setattr(
+        ic.InfracostClient, "sync_to_cache",
+        lambda self, cache, usage_metric, region, vendor="aws":
+            calls.append(region) or 1,
+    )
+    ic.sync_pricing_catalog(services=["KMS-Key-Month"])
+    assert calls == ["us-east-1"]