unique-toolkit 2026.32.0.dev4__tar.gz → 2026.32.0.dev5__tar.gz

{unique_toolkit-2026.32.0.dev4 → unique_toolkit-2026.32.0.dev5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: unique-toolkit
-Version: 2026.32.0.dev4
+Version: 2026.32.0.dev5
 Summary: 
 Author: Cedric Klinkert, Martin Fadler, Sadique Sheik, Fabian Schläpfer, Pascal Hauri, Ejiroghene Oki-Peter, Rami Azouz, Ricardo Visini, Sanzio Monti
 Author-email: Cedric Klinkert <cedric.klinkert@unique.ch>, Martin Fadler <martin.fadler@unique.ch>, Sadique Sheik <sadique@unique.ch>, Fabian Schläpfer <fabian@unique.ch>, Pascal Hauri <pascal@unique.ch>, Ejiroghene Oki-Peter <oki-peter.ejiroghene@unique.ai>, Rami Azouz <rami.ext@unique.ch>, Ricardo Visini <ricardo.visini@unique.ch>, Sanzio Monti <sanzio.monti@unique.ai>
@@ -16,7 +16,7 @@ Requires-Dist: sseclient>=0.0.27,<0.0.28
 Requires-Dist: openai>=1.105.0,<3
 Requires-Dist: platformdirs>=4.0.0,<5
 Requires-Dist: pillow>=12.1.1
-Requires-Dist: unique-sdk>=2026.32.0.dev5,<2026.32.0rc0
+Requires-Dist: unique-sdk>=2026.32.0.dev8,<2026.32.0rc0
 Requires-Dist: jambo>=0.1.2,<0.2
 Requires-Dist: jinja2>=3.1.6,<4
 Requires-Dist: docxtpl>=0.20.1,<0.21

{unique_toolkit-2026.32.0.dev4 → unique_toolkit-2026.32.0.dev5}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "unique_toolkit"
-version = "2026.32.0.dev4"
+version = "2026.32.0.dev5"
 description = ""
 readme = "README.md"
 requires-python = ">=3.12"
@@ -28,7 +28,7 @@ dependencies = [
     "openai>=1.105.0,<3",
     "platformdirs>=4.0.0,<5",
     "pillow>=12.1.1",
-    "unique-sdk>=2026.32.0.dev5,<2026.32.0rc0",
+    "unique-sdk>=2026.32.0.dev8,<2026.32.0rc0",
     "jambo>=0.1.2,<0.2",
     "jinja2>=3.1.6,<4",
     "docxtpl>=0.20.1,<0.21",

{unique_toolkit-2026.32.0.dev4 → unique_toolkit-2026.32.0.dev5}/unique_toolkit/monitoring/middleware.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import time
+from collections.abc import Sequence
 
 from prometheus_client import (  # pyright: ignore[reportMissingImports]
     Counter,
@@ -16,12 +17,6 @@ _http_requests_total = Counter(
     ["method", "path", "status_code"],
     registry=REGISTRY,
 )
-_http_request_duration_seconds = Histogram(
-    "python_http_request_duration_seconds",
-    "HTTP request duration",
-    ["method", "path"],
-    registry=REGISTRY,
-)
 _http_requests_in_progress = Gauge(
     "python_http_requests_in_progress",
     "In-flight HTTP requests",
@@ -30,6 +25,35 @@ _http_requests_in_progress = Gauge(
     multiprocess_mode="livesum",
 )
 
+# Created lazily on first MetricsMiddleware construction so services can choose the
+# bucket layout. Prometheus allows a metric name to be registered only once per
+# registry, so the buckets passed to the first middleware win for the process.
+_http_request_duration_seconds: Histogram | None = None
+
+
+def _get_or_create_duration_histogram(
+    buckets: Sequence[float] | None,
+) -> Histogram:
+    global _http_request_duration_seconds
+    if _http_request_duration_seconds is None:
+        if buckets is None:
+            _http_request_duration_seconds = Histogram(
+                "python_http_request_duration_seconds",
+                "HTTP request duration",
+                ["method", "path"],
+                registry=REGISTRY,
+            )
+        else:
+            _http_request_duration_seconds = Histogram(
+                "python_http_request_duration_seconds",
+                "HTTP request duration",
+                ["method", "path"],
+                registry=REGISTRY,
+                buckets=tuple(buckets),
+            )
+    return _http_request_duration_seconds
+
+
 _DEFAULT_EXCLUDED_PATHS = frozenset({"/health", "/metrics", "/"})
 
 
@@ -51,13 +75,29 @@ class MetricsMiddleware:
 
         # Quart
         app.asgi_app = MetricsMiddleware(app.asgi_app)
+
+    ``duration_buckets`` overrides the latency histogram buckets (seconds). Without
+    it, prometheus_client's defaults apply and their top finite bucket is 10s —
+    histogram_quantile cannot resolve past the largest finite bucket, so services
+    with slower endpoints should pass buckets covering their full latency range::
+
+        app.add_middleware(MetricsMiddleware, duration_buckets=(1, 5, 30, 60, 120))
+
+    The histogram is registered once per process; the buckets of the first
+    constructed middleware apply.
     """
 
-    def __init__(self, app, excluded_paths: set[str] | None = None):
+    def __init__(
+        self,
+        app,
+        excluded_paths: set[str] | None = None,
+        duration_buckets: Sequence[float] | None = None,
+    ):
         self.app = app
         self._excluded_paths = (
             excluded_paths if excluded_paths is not None else _DEFAULT_EXCLUDED_PATHS
         )
+        self._duration = _get_or_create_duration_histogram(duration_buckets)
 
     async def __call__(self, scope, receive, send):
         # Pass through non-HTTP (WebSocket, lifespan)
@@ -87,9 +127,7 @@ class MetricsMiddleware:
         finally:
             duration = time.perf_counter() - start
             _http_requests_in_progress.labels(method=method).dec()
-            _http_request_duration_seconds.labels(method=method, path=path).observe(
-                duration
-            )
+            self._duration.labels(method=method, path=path).observe(duration)
             _http_requests_total.labels(
                 method=method, path=path, status_code=status_code
             ).inc()