diff --git a/lib/src/sql/idiom.rs b/lib/src/sql/idiom.rs
index 40921600..aa40210c 100644
--- a/lib/src/sql/idiom.rs
+++ b/lib/src/sql/idiom.rs
@@ -130,6 +130,12 @@ impl Idiom {
 	pub(crate) fn split_multi_yield(v: &Part) -> bool {
 		matches!(v, Part::Graph(g) if g.alias.is_some())
 	}
+	/// Check if the path part is a yield in a multi-yield expression
+	pub(crate) fn remove_trailing_all(&mut self) {
+		if self.ends_with(&[Part::All]) {
+			self.0.truncate(self.len() - 1);
+		}
+	}
 }
 
 impl Idiom {
diff --git a/lib/src/sql/value/every.rs b/lib/src/sql/value/every.rs
index e939ddfb..b66c162b 100644
--- a/lib/src/sql/value/every.rs
+++ b/lib/src/sql/value/every.rs
@@ -9,39 +9,49 @@ impl Value {
 			None => self._every(steps, arrays, Idiom::default()),
 		}
 	}
-	fn _every(&self, steps: bool, arrays: bool, prev: Idiom) -> Vec<Idiom> {
+	fn _every(&self, steps: bool, arrays: bool, mut prev: Idiom) -> Vec<Idiom> {
 		match self {
 			// Current path part is an object and is not empty
-			Value::Object(v) if !v.is_empty() => match steps {
-				// Let's log all intermediary nodes
-				true if !prev.is_empty() => Some(prev.clone())
-					.into_iter()
-					.chain(v.iter().flat_map(|(k, v)| {
-						let p = Part::from(k.to_owned());
-						v._every(steps, arrays, prev.clone().push(p))
-					}))
-					.collect::<Vec<_>>(),
-				// Let's not log intermediary nodes
-				_ => v
-					.iter()
-					.flat_map(|(k, v)| {
-						let p = Part::from(k.to_owned());
-						v._every(steps, arrays, prev.clone().push(p))
-					})
-					.collect::<Vec<_>>(),
-			},
+			Value::Object(v) if !v.is_empty() => {
+				// Remove any trailing * path parts
+				prev.remove_trailing_all();
+				// Check if we should log intermediary nodes
+				match steps {
+					// Let's log all intermediary nodes
+					true if !prev.is_empty() => Some(prev.clone())
+						.into_iter()
+						.chain(v.iter().flat_map(|(k, v)| {
+							let p = Part::from(k.to_owned());
+							v._every(steps, arrays, prev.clone().push(p))
+						}))
+						.collect::<Vec<_>>(),
+					// Let's not log intermediary nodes
+					_ => v
+						.iter()
+						.flat_map(|(k, v)| {
+							let p = Part::from(k.to_owned());
+							v._every(steps, arrays, prev.clone().push(p))
+						})
+						.collect::<Vec<_>>(),
+				}
+			}
 			// Current path part is an array and is not empty
-			Value::Array(v) if !v.is_empty() => match arrays {
-				// Let's log all individual array items
-				true => std::iter::once(prev.clone())
-					.chain(v.iter().enumerate().rev().flat_map(|(i, v)| {
-						let p = Part::from(i.to_owned());
-						v._every(steps, arrays, prev.clone().push(p))
-					}))
-					.collect::<Vec<_>>(),
-				// Let's not log individual array items
-				false => vec![prev],
-			},
+			Value::Array(v) if !v.is_empty() => {
+				// Remove any trailing * path parts
+				prev.remove_trailing_all();
+				// Check if we should log individual array items
+				match arrays {
+					// Let's log all individual array items
+					true => std::iter::once(prev.clone())
+						.chain(v.iter().enumerate().rev().flat_map(|(i, v)| {
+							let p = Part::from(i.to_owned());
+							v._every(steps, arrays, prev.clone().push(p))
+						}))
+						.collect::<Vec<_>>(),
+					// Let's not log individual array items
+					false => vec![prev],
+				}
+			}
 			// Process everything else
 			_ => vec![prev],
 		}
@@ -117,4 +127,23 @@ mod tests {
 		];
 		assert_eq!(res, val.every(None, true, true));
 	}
+
+	#[test]
+	fn every_including_intermediary_nodes_including_array_indexes_ending_all() {
+		let val = Value::parse("{ test: { something: [{ age: 34, tags: ['code', 'databases'] }, { age: 36, tags: ['design', 'operations'] }] } }");
+		let res = vec![
+			Idiom::parse("test.something"),
+			Idiom::parse("test.something[1]"),
+			Idiom::parse("test.something[1].age"),
+			Idiom::parse("test.something[1].tags"),
+			Idiom::parse("test.something[1].tags[1]"),
+			Idiom::parse("test.something[1].tags[0]"),
+			Idiom::parse("test.something[0]"),
+			Idiom::parse("test.something[0].age"),
+			Idiom::parse("test.something[0].tags"),
+			Idiom::parse("test.something[0].tags[1]"),
+			Idiom::parse("test.something[0].tags[0]"),
+		];
+		assert_eq!(res, val.every(Some(&Idiom::parse("test.something.*")), true, true));
+	}
 }
diff --git a/lib/tests/field.rs b/lib/tests/field.rs
index 25631964..1c45b000 100644
--- a/lib/tests/field.rs
+++ b/lib/tests/field.rs
@@ -920,3 +920,73 @@ async fn field_definition_readonly() -> Result<(), Error> {
 	//
 	Ok(())
 }
+
+#[tokio::test]
+async fn field_definition_flexible_array_any() -> Result<(), Error> {
+	let sql = "
+		DEFINE TABLE user SCHEMAFULL;
+		DEFINE FIELD custom ON user TYPE option<array>;
+		DEFINE FIELD custom.* ON user FLEXIBLE TYPE any;
+		CREATE user:one CONTENT { custom: ['sometext'] };
+		CREATE user:two CONTENT { custom: [ ['sometext'] ] };
+		CREATE user:three CONTENT { custom: [ { key: 'sometext' } ] };
+	";
+	let dbs = new_ds().await?.with_auth_enabled(true);
+	let ses = Session::owner().with_ns("test").with_db("test");
+	let res = &mut dbs.execute(sql, &ses, None).await?;
+	assert_eq!(res.len(), 6);
+	//
+	let tmp = res.remove(0).result;
+	assert!(tmp.is_ok());
+	//
+	let tmp = res.remove(0).result;
+	assert!(tmp.is_ok());
+	//
+	let tmp = res.remove(0).result;
+	assert!(tmp.is_ok());
+	//
+	let tmp = res.remove(0).result?;
+	let val = Value::parse(
+		"[
+			{
+				custom: [
+					'sometext'
+				],
+				id: user:one
+			},
+		]",
+	);
+	assert_eq!(tmp, val);
+	//
+	let tmp = res.remove(0).result?;
+	let val = Value::parse(
+		"[
+			{
+				custom: [
+					[
+						'sometext'
+					]
+				],
+				id: user:two
+			},
+		]",
+	);
+	assert_eq!(tmp, val);
+	//
+	let tmp = res.remove(0).result?;
+	let val = Value::parse(
+		"[
+			{
+				custom: [
+					{
+						key: 'sometext'
+					}
+				],
+				id: user:three
+			}
+		]",
+	);
+	assert_eq!(tmp, val);
+	//
+	Ok(())
+}