chore: import upstream snapshot with attribution

.github/ISSUE_TEMPLATE/feature_request.yml

@@ -0,0 +1,55 @@
+name: Feature Request
+description: Suggest a new feature or improvement (e.g., better memory control, new query option)
+title: "[Feature]: "
+labels: ["feature"]
+body:
+  - type: markdown
+    attributes:
+      value: |
+        Thanks for your idea! Help us understand the motivation and scope.
+
+  - type: textarea
+    id: problem_or_motivation
+    attributes:
+      label: Problem / Motivation
+      description: What problem does this solve? Why is it needed?
+      placeholder: |
+        e.g. "Current vector queries don't allow filtering by metadata + distance threshold at once"
+    validations:
+      required: true
+
+  - type: textarea
+    id: proposed_solution
+    attributes:
+      label: Proposed Solution
+      description: How should it work? API sketch or pseudocode welcome.
+      placeholder: |
+        Add `filter=` and `max_distance=` args to `Zvec.query()`:
+        ```python
+        results = db.query(vector, filter="category == 'A'", max_distance=0.5)
+        ```
+      render: python
+    validations:
+      required: false
+
+  - type: textarea
+    id: alternatives
+    attributes:
+      label: Alternatives Considered
+      description: Are there workarounds? Why not use them?
+    validations:
+      required: false
+
+  - type: dropdown
+    id: impact_area
+    attributes:
+      label: Affected Area
+      multiple: true
+      options:
+        - label: C++ Core (storage, indexing)
+        - label: Python API / Bindings
+        - label: Build System (CMake, Homebrew pkg)
+        - label: Testing / CI / Coverage
+        - label: Documentation
+    validations:
+      required: false