gooddata-eval 1.68.1.dev3__py3-none-any.whl → 1.68.1.dev4__py3-none-any.whl

gooddata_eval/core/agentic/alert_skill.py

@@ -77,6 +77,8 @@ def _check_threshold(expected: CatalogMetricAlert, actual_args: dict) -> bool:
 
 
 def _check_trigger(expected: CatalogMetricAlert, actual_args: dict) -> bool:
+    if expected.operator == "ANOMALY":
+        return True
     exp_trigger = expected.trigger
     act_trigger = actual_args.get("trigger", actual_args.get("triggerMode", "ALWAYS"))
     if exp_trigger in _ALWAYS_TRIGGER_VALUES:
@@ -235,27 +237,42 @@ class AgenticAlertSummary:
     best: AlertRunResult
 
 
+def _case_insensitive_get(d: dict, *keys: str) -> Any:
+    """Look up a value by key, preferring an exact match then a case-insensitive one."""
+    for k in keys:
+        if k in d:
+            return d[k]
+    lowered = {str(k).lower(): v for k, v in d.items()}
+    for k in keys:
+        if k.lower() in lowered:
+            return lowered[k.lower()]
+    return None
+
+
 def _normalize_expected_output(expected: dict) -> CatalogMetricAlert:
     """Parse expected_output dict into CatalogMetricAlert, accepting display-format or internal-format keys."""
-    operator = expected.get("operator") or expected.get("Operator") or "GREATER_THAN"
-    threshold = expected.get("threshold") or expected.get("Threshold")
-    threshold_from = expected.get("threshold_from")
-    threshold_to = expected.get("threshold_to")
-    trigger = expected.get("trigger") or expected.get("Trigger") or "not specified"
-
-    metric_id = expected.get("metric_id")
-    if not metric_id and "Metric" in expected:
-        m = re.search(r"\(([^)]+)\)\s*$", str(expected["Metric"]))
-        if m:
-            metric_id = m.group(1).strip()
-
-    raw_recip = expected.get("recipients") or expected.get("Recipient(s)") or []
+    operator = _case_insensitive_get(expected, "operator") or "GREATER_THAN"
+    threshold = _case_insensitive_get(expected, "threshold")
+    threshold_from = _case_insensitive_get(expected, "threshold_from")
+    threshold_to = _case_insensitive_get(expected, "threshold_to")
+
+    trigger = _case_insensitive_get(expected, "trigger") or "not specified"
+    trigger = _TRIGGER_DISPLAY_TO_API.get(trigger, trigger)
+
+    metric_id = _case_insensitive_get(expected, "metric_id")
+    if not metric_id:
+        metric_disp = _case_insensitive_get(expected, "metric")
+        if metric_disp:
+            m = re.search(r"\(([^)]+)\)\s*$", str(metric_disp))
+            metric_id = m.group(1).strip() if m else None
+
+    raw_recip = _case_insensitive_get(expected, "recipients", "recipient(s)") or []
     if isinstance(raw_recip, str):
         recipients = [r.strip() for r in raw_recip.replace(";", ",").split(",") if r.strip()]
     else:
         recipients = list(raw_recip)
 
-    filters = expected.get("filters") or expected.get("Time window/Filters")
+    filters = _case_insensitive_get(expected, "filters")
     if isinstance(filters, str) and any(kw in filters for kw in ("None", "All time")):
         filters = None
 

gooddata_eval/core/agentic/conversation.py

@@ -102,7 +102,7 @@ def _activated_skills(tool_call_events: list[ToolCallEvent]) -> list[str]:
         if tc.function_name != "set_skills":
             continue
         args = tc.parsed_arguments() or {}
-        skills.extend(args.get("skills", []))
+        skills.extend(args.get("skill_names") or args.get("skills") or [])
     return list(set(skills))
 
 

gooddata_eval/core/scoring.py

@@ -71,7 +71,9 @@ def validate_cross_references(viz: CreatedVisualization) -> tuple[bool, list[str
             continue
         using_val = filter_dict.get("using", "")
         using_uri = _resolve_alias_to_uri(using_val, fields)
-        if not using_uri.startswith(("metric/", "fact/")):
+        field_def = fields.get(using_val)
+        is_adhoc_agg = isinstance(field_def, AacQueryField) and bool(field_def.aggregation)
+        if not using_uri.startswith(("metric/", "fact/")) and not is_adhoc_agg:
             errors.append(
                 f"ranking filter '{filter_key}': using='{using_val}' "
                 f"resolves to '{using_uri}' — expected a metric/ or fact/ URI"

{gooddata_eval-1.68.1.dev3.dist-info → gooddata_eval-1.68.1.dev4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gooddata-eval
-Version: 1.68.1.dev3
+Version: 1.68.1.dev4
 Summary: Evaluate the GoodData AI agent against your own questions and models.
 Project-URL: Source, https://github.com/gooddata/gooddata-python-sdk
 Author-email: GoodData <support@gooddata.com>
@@ -17,7 +17,7 @@ Classifier: Topic :: Scientific/Engineering
 Classifier: Topic :: Software Development
 Classifier: Typing :: Typed
 Requires-Python: >=3.10
-Requires-Dist: gooddata-sdk~=1.68.1.dev3
+Requires-Dist: gooddata-sdk~=1.68.1.dev4
 Requires-Dist: httpx<1.0,>=0.27
 Requires-Dist: orjson<4.0.0,>=3.9.15
 Requires-Dist: pydantic<3.0,>=2.6

{gooddata_eval-1.68.1.dev3.dist-info → gooddata_eval-1.68.1.dev4.dist-info}/RECORD

@@ -8,13 +8,13 @@ gooddata_eval/core/config.py,sha256=c0lgsILU70VJejhJGSZJmyne8i2uNXDiZEXRRKJEOgs,
 gooddata_eval/core/connection.py,sha256=mh-Do5-Ti0-DXXikfCIH10eTTUbKuZUAS_sLoPNDJ5I,1229
 gooddata_eval/core/models.py,sha256=g7bi7OGxnADGidVQmJLNuvehddSoW_-4e4YhEFSSkKs,3989
 gooddata_eval/core/runner.py,sha256=vRo2SwRKRjkSNCp6tse4tSGRWC3zBPFdjAnM5rMg05M,8064
-gooddata_eval/core/scoring.py,sha256=eGn72OuPc_Hnm6oNmoWkr-HzFOJP06Yay7BRy6BEN9Q,5852
+gooddata_eval/core/scoring.py,sha256=7ku48WEXtZl8EhowKikMiBmKfH7BDoZxHVvEOY8h3Rk,6007
 gooddata_eval/core/workspace.py,sha256=S8exiy5gpDtcRtT8-i2fIRALF34Pbo0J46xm8X_ldus,11846
 gooddata_eval/core/agentic/__init__.py,sha256=nPjJ9sbnJVvsAlo9QMQlnPFgLNTZRzy9-0IKlnRcU8Q,2671
 gooddata_eval/core/agentic/_catalog.py,sha256=eCaeyvLF-FCN0J3UnXSHaKtufU6HvrM2mxnmgyXsY1U,2085
 gooddata_eval/core/agentic/_langfuse.py,sha256=zJiZTz5xloWQBunwxXzdOY3NrA0S4QPFc6d3iB-cgkE,13432
-gooddata_eval/core/agentic/alert_skill.py,sha256=dOHEnE3823P4tx0FgoPcdPcu-i-qpju20LZFR6N8pBU,19138
-gooddata_eval/core/agentic/conversation.py,sha256=Ynq9Rg6vVXjpHRe76r34-KPISp-N9TZxm9rk2cNcibg,17513
+gooddata_eval/core/agentic/alert_skill.py,sha256=h5i-GVJOAUzgMhIF5yGKIz7JL3f6QXKZAKz6gcN7vpY,19702
+gooddata_eval/core/agentic/conversation.py,sha256=tTpN1AySPGDbYcpwY2G0J38lRYiU5sqvbdDJBOyPQHc,17542
 gooddata_eval/core/agentic/general_question.py,sha256=piPY5XOMjH4Xwl2qN0D5KWBL1RJjYaP63oX1eXp-7yU,7978
 gooddata_eval/core/agentic/guardrail.py,sha256=-vZg-jeL2LYKGaLesuu7Yg1AIEn-9bfDK130EMdw_7s,7865
 gooddata_eval/core/agentic/metric_skill.py,sha256=jFqVsxNEx-7IHa4v8aNLkDzD9OogDsgV4O3UJkiVDFw,11388
@@ -44,8 +44,8 @@ gooddata_eval/core/reporting/console.py,sha256=kCHt4sCSaSRh0ZWqaSEJOl6_Jgb6uiouS
 gooddata_eval/core/reporting/json_report.py,sha256=NIA_PvnZ0PbJnXz_zZauxc1gGnzqlLD_ouVqWeNmjtw,2837
 gooddata_eval/core/summary/__init__.py,sha256=U54lANKm34yjqm7dZL9KkGouPbwAaQWC9wiAmb5B5g4,32
 gooddata_eval/core/summary/http_client.py,sha256=dPArJ6zoCu1W9xmYXFA1WMDNsSOkhn3gaCpYgCUp3gk,2174
-gooddata_eval-1.68.1.dev3.dist-info/METADATA,sha256=ZGte7olPAbOw-BM69NnUhCskaAP4lSQRwwFTnapqJJM,9513
-gooddata_eval-1.68.1.dev3.dist-info/WHEEL,sha256=mffPy8wBnZQn2VnJUU5jE99KsxaSfiyMHV9Yt0aLVxs,87
-gooddata_eval-1.68.1.dev3.dist-info/entry_points.txt,sha256=28nFp5Viknx4haPYgzK9QHlwPFfC6rPF6RmjC2ht8EI,56
-gooddata_eval-1.68.1.dev3.dist-info/licenses/LICENSE.txt,sha256=LVfVlC9maU3K9lgMwdZnClQQsOIbOekdcdVKr9qzJtk,256836
-gooddata_eval-1.68.1.dev3.dist-info/RECORD,,
+gooddata_eval-1.68.1.dev4.dist-info/METADATA,sha256=7eLGgge0gSIxqwAWeU3kPoPa4VjLC-WH0tQne-KWbS8,9513
+gooddata_eval-1.68.1.dev4.dist-info/WHEEL,sha256=mffPy8wBnZQn2VnJUU5jE99KsxaSfiyMHV9Yt0aLVxs,87
+gooddata_eval-1.68.1.dev4.dist-info/entry_points.txt,sha256=28nFp5Viknx4haPYgzK9QHlwPFfC6rPF6RmjC2ht8EI,56
+gooddata_eval-1.68.1.dev4.dist-info/licenses/LICENSE.txt,sha256=LVfVlC9maU3K9lgMwdZnClQQsOIbOekdcdVKr9qzJtk,256836
+gooddata_eval-1.68.1.dev4.dist-info/RECORD,,