diff --git a/lib/galaxy/model/store/discover.py b/lib/galaxy/model/store/discover.py
index 679bd2aa38f7..17389c7b83eb 100644
--- a/lib/galaxy/model/store/discover.py
+++ b/lib/galaxy/model/store/discover.py
@@ -38,6 +38,10 @@
 from galaxy.util.hash_util import HASH_NAME_MAP
 
 if TYPE_CHECKING:
+    from galaxy.job_execution.output_collect import (
+        DatasetCollector,
+        ToolMetadataDatasetCollector,
+    )
     from galaxy.model.store import ModelExportStore
 
 log = logging.getLogger(__name__)
@@ -50,7 +54,7 @@ class MaxDiscoveredFilesExceededError(ValueError):
     pass
 
 
-CollectorT = Any  # TODO: setup an interface for these file collectors data classes.
+CollectorT = Union["DatasetCollector", "ToolMetadataDatasetCollector"]
 
 
 class ModelPersistenceContext(metaclass=abc.ABCMeta):
@@ -1056,19 +1060,21 @@ def name(self):
         return self.as_dict.get("name")
 
     @property
-    def dbkey(self):
-        return self.as_dict.get("dbkey", getattr(self.collector, "default_dbkey", "?"))
+    def dbkey(self) -> str:
+        return self.as_dict.get("dbkey", self.collector and self.collector.default_dbkey or "?")
 
     @property
-    def ext(self):
-        return self.as_dict.get("ext", getattr(self.collector, "default_ext", "data"))
+    def ext(self) -> str:
+        return self.as_dict.get("ext", self.collector and self.collector.default_ext or "data")
 
     @property
-    def visible(self):
+    def visible(self) -> bool:
         try:
             return self.as_dict["visible"].lower() == "visible"
         except KeyError:
-            return getattr(self.collector, "default_visible", True)
+            if self.collector and self.collector.default_visible is not None:
+                return self.collector.default_visible
+            return True
 
     @property
     def link_data(self):
diff --git a/lib/galaxy/tool_util/parser/output_collection_def.py b/lib/galaxy/tool_util/parser/output_collection_def.py
index 4d16cb5d3539..b7929ebe1665 100644
--- a/lib/galaxy/tool_util/parser/output_collection_def.py
+++ b/lib/galaxy/tool_util/parser/output_collection_def.py
@@ -34,7 +34,13 @@ def dataset_collector_descriptions_from_elem(elem, legacy=True):
     if num_discover_dataset_blocks == 0 and legacy:
         collectors = [DEFAULT_DATASET_COLLECTOR_DESCRIPTION]
     else:
-        collectors = [dataset_collection_description(**e.attrib) for e in primary_dataset_elems]
+        default_format = elem.attrib.get("format")
+        collectors = []
+        for e in primary_dataset_elems:
+            description_attributes = e.attrib
+            if default_format and "format" not in description_attributes and "ext" not in description_attributes:
+                description_attributes["format"] = default_format
+            collectors.append(dataset_collection_description(**description_attributes))
 
     return _validate_collectors(collectors)
 
diff --git a/lib/galaxy/tool_util/xsd/galaxy.xsd b/lib/galaxy/tool_util/xsd/galaxy.xsd
index bc1f18780a60..daebc29070ac 100644
--- a/lib/galaxy/tool_util/xsd/galaxy.xsd
+++ b/lib/galaxy/tool_util/xsd/galaxy.xsd
@@ -5360,11 +5360,13 @@ The default is ``galaxy.json``.
   <xs:attributeGroup name="OutputCommon">
     <xs:attribute name="format" type="xs:string">
       <xs:annotation>
-        <xs:documentation xml:lang="en">The short name for the output datatype.
-The valid values for format can be found in
+        <xs:documentation xml:lang="en"><![CDATA[
+The short name for the output datatype. The valid values for format can be found in
 [/config/datatypes_conf.xml.sample](https://github.com/galaxyproject/galaxy/blob/dev/config/datatypes_conf.xml.sample)
-(e.g. ``format="pdf"`` or ``format="fastqsanger"``). For collections this is the default format for all included
-elements. Note that the format specified here is ignored for discovered data sets.</xs:documentation>
+(e.g. ``format="pdf"`` or ``format="fastqsanger"``). For collections this is the default
+format for all included elements. Note that the format specified here is ignored for
+discovered data sets on Galaxy versions prior to 24.0 and should be specified using the ``<discovered_data>`` tag set.
+        ]]></xs:documentation>
       </xs:annotation>
     </xs:attribute>
     <xs:attribute name="format_source" type="xs:string">
diff --git a/test/functional/tools/discover_default_ext.xml b/test/functional/tools/discover_default_ext.xml
new file mode 100644
index 000000000000..7f7c7d83cdd8
--- /dev/null
+++ b/test/functional/tools/discover_default_ext.xml
@@ -0,0 +1,52 @@
+<tool id="discover_default_ext" name="discover_default_ext" version="0.1.0">
+    <command><![CDATA[
+echo 1 > 1.txt;
+]]></command>
+    <inputs />
+    <outputs>
+        <collection name="collection_with_default_ext" type="list" label="with default format" format="fasta">
+            <discover_datasets pattern="__name_and_ext__" />
+        </collection>
+        <collection name="collection_default_ext_and_explicit_format" type="list" label="with default format and static element format" format="fasta">
+            <discover_datasets pattern="__name__" format="txt" />
+        </collection>
+        <collection name="collection_default_ext_used" type="list" label="with default format and no override" format="fasta">
+            <discover_datasets pattern="__name__" />
+        </collection>
+        <collection name="collection_without_default_ext" type="list" label="wihtout default ext, should be data">
+            <discover_datasets pattern="__name__" />
+        </collection>
+    </outputs>
+    <tests>
+        <test expect_num_outputs="4">
+            <output_collection name="collection_with_default_ext" type="list" count="1">
+                <element name="1" ftype="txt">
+                    <assert_contents>
+                        <has_text text="1" />
+                    </assert_contents>
+                </element>
+            </output_collection>
+            <output_collection name="collection_default_ext_and_explicit_format" type="list" count="1">
+                <element name="1.txt" ftype="txt">
+                    <assert_contents>
+                        <has_text text="1" />
+                    </assert_contents>
+                </element>
+            </output_collection>
+            <output_collection name="collection_default_ext_used" type="list" count="1">
+                <element name="1.txt" ftype="fasta">
+                    <assert_contents>
+                        <has_text text="1" />
+                    </assert_contents>
+                </element>
+            </output_collection>
+            <output_collection name="collection_without_default_ext" type="list" count="1">
+                <element name="1.txt" ftype="data">
+                    <assert_contents>
+                        <has_text text="1" />
+                    </assert_contents>
+                </element>
+            </output_collection>
+        </test>
+    </tests>
+</tool>
diff --git a/test/functional/tools/sample_tool_conf.xml b/test/functional/tools/sample_tool_conf.xml
index 2c074381db5e..a9277f4fab78 100644
--- a/test/functional/tools/sample_tool_conf.xml
+++ b/test/functional/tools/sample_tool_conf.xml
@@ -206,6 +206,7 @@
   <tool file="collection_cat_group_tag.xml" />
   <tool file="collection_cat_group_tag_multiple.xml" />
   <tool file="discover_sort_by.xml" />
+  <tool file="discover_default_ext.xml" />
   <tool file="expression_forty_two.xml" />
   <tool file="expression_pick_larger_file.xml" />
   <tool file="expression_parse_int.xml" />
diff --git a/test/unit/app/tools/test_collect_primary_datasets.py b/test/unit/app/tools/test_collect_primary_datasets.py
index 16f3c6aae13e..fb612853b4db 100644
--- a/test/unit/app/tools/test_collect_primary_datasets.py
+++ b/test/unit/app/tools/test_collect_primary_datasets.py
@@ -128,6 +128,20 @@ def test_collect_multiple_recurse_dict(self):
         created_hda_3 = datasets[DEFAULT_TOOL_OUTPUT]["test3"]
         assert_created_with_path(self.app.object_store, created_hda_3.dataset, path3)
 
+    def test_collect_collection_default_format(self):
+        self._replace_output_collectors(
+            """<dataset_collection name="parent" format="abcdef">
+            <discover_datasets pattern="__name__" directory="subdir_for_name_discovery" sort_by="reverse_filename" />
+        </dataset_collection>"""
+        )
+        self._setup_extra_file(subdir="subdir_for_name_discovery", filename="test1")
+        self._setup_extra_file(subdir="subdir_for_name_discovery", filename="test2")
+
+        datasets = self._collect()
+        assert DEFAULT_TOOL_OUTPUT in datasets
+        for dataset in datasets[DEFAULT_TOOL_OUTPUT].values():
+            assert dataset.ext == "abcdef"
+
     def test_collect_sorted_reverse(self):
         self._replace_output_collectors(
             """<output>