diff --git a/src/lean_spec/subspecs/xmss/aggregation.py b/src/lean_spec/subspecs/xmss/aggregation.py
index 4df52745..6b2ba009 100644
--- a/src/lean_spec/subspecs/xmss/aggregation.py
+++ b/src/lean_spec/subspecs/xmss/aggregation.py
@@ -7,8 +7,6 @@
 - Type-2: a merge of N Type-1 proofs over distinct messages.
 """
 
-from __future__ import annotations
-
 from collections.abc import Sequence
 
 from lean_multisig_py import (
@@ -67,8 +65,8 @@ class TypeOneMultiSignature(Container):
 
     @staticmethod
     def select_greedily(
-        *proof_sets: set[TypeOneMultiSignature] | None,
-    ) -> tuple[list[TypeOneMultiSignature], set[ValidatorIndex]]:
+        *proof_sets: set["TypeOneMultiSignature"] | None,
+    ) -> tuple[list["TypeOneMultiSignature"], set[ValidatorIndex]]:
         """Greedy set-cover over Type-1 proofs to maximise validator coverage.
 
         Repeatedly selects the proof covering the most uncovered validators
@@ -102,13 +100,13 @@ def select_greedily(
 
     @staticmethod
     def aggregate(
-        children: Sequence[tuple[TypeOneMultiSignature, Sequence[PublicKey]]],
+        children: Sequence[tuple["TypeOneMultiSignature", Sequence[PublicKey]]],
         raw_xmss: Sequence[tuple[PublicKey, Signature]],
         xmss_participants: AggregationBits | None,
         message: Bytes32,
         slot: Slot,
         mode: LeanEnvMode | None = None,
-    ) -> TypeOneMultiSignature:
+    ) -> "TypeOneMultiSignature":
         """Aggregate raw XMSS signatures and child Type-1 proofs into one Type-1 proof.
 
         Proof bytes are stored in compact no-pubkeys form. Participant identity is
@@ -224,7 +222,7 @@ def aggregate(
         parts: Sequence[TypeOneMultiSignature],
         public_keys_per_part: Sequence[Sequence[PublicKey]] | None = None,
         mode: LeanEnvMode | None = None,
-    ) -> TypeTwoMultiSignature:
+    ) -> "TypeTwoMultiSignature":
         """Merge several Type-1 proofs (each over a distinct message) into one Type-2 proof.
 
         The returned Type-2 proof bytes are stored in compact no-pubkeys form.
diff --git a/src/lean_spec/subspecs/xmss/containers.py b/src/lean_spec/subspecs/xmss/containers.py
index 07198677..02ec1f85 100644
--- a/src/lean_spec/subspecs/xmss/containers.py
+++ b/src/lean_spec/subspecs/xmss/containers.py
@@ -1,7 +1,5 @@
 """Generalized XMSS containers."""
 
-from __future__ import annotations
-
 from typing import override
 
 from pydantic import field_serializer, field_validator, model_serializer
diff --git a/src/lean_spec/subspecs/xmss/subtree.py b/src/lean_spec/subspecs/xmss/subtree.py
index 22d1f973..973e6e2e 100644
--- a/src/lean_spec/subspecs/xmss/subtree.py
+++ b/src/lean_spec/subspecs/xmss/subtree.py
@@ -5,7 +5,7 @@
 implementing the memory-efficient top-bottom tree traversal approach.
 """
 
-from __future__ import annotations
+from typing import Self
 
 from lean_spec.types import Uint64
 from lean_spec.types.container import Container
@@ -103,7 +103,7 @@ def new(
         start_index: Uint64,
         parameter: Parameter,
         lowest_layer_nodes: list[HashDigestVector],
-    ) -> HashSubTree:
+    ) -> Self:
         """
         Builds a new sparse Merkle subtree starting from a specified layer.
 
@@ -187,7 +187,7 @@ def new_top_tree(
         start_bottom_tree_index: Uint64,
         parameter: Parameter,
         bottom_tree_roots: list[HashDigestVector],
-    ) -> HashSubTree:
+    ) -> Self:
         """
         Constructs a top tree from the roots of bottom trees.
 
@@ -243,7 +243,7 @@ def new_bottom_tree(
         bottom_tree_index: Uint64,
         parameter: Parameter,
         leaves: list[HashDigestVector],
-    ) -> HashSubTree:
+    ) -> Self:
         """
         Constructs a single bottom tree from leaf hashes.
 
@@ -327,7 +327,7 @@ def from_prf_key(
         prf_key: PRFKey,
         bottom_tree_index: Uint64,
         parameter: Parameter,
-    ) -> HashSubTree:
+    ) -> Self:
         """
         Generates a single bottom tree on-demand from the PRF key.
 
diff --git a/src/lean_spec/types/bitfields.py b/src/lean_spec/types/bitfields.py
index 0c2f2c07..169c1910 100644
--- a/src/lean_spec/types/bitfields.py
+++ b/src/lean_spec/types/bitfields.py
@@ -12,8 +12,6 @@
 Bit i of the input lands in byte i // 8 at position i % 8.
 """
 
-from __future__ import annotations
-
 import math
 from collections.abc import Sequence
 from typing import (
diff --git a/src/lean_spec/types/byte_arrays.py b/src/lean_spec/types/byte_arrays.py
index 82f19758..c59b2747 100644
--- a/src/lean_spec/types/byte_arrays.py
+++ b/src/lean_spec/types/byte_arrays.py
@@ -11,8 +11,6 @@
 Both flavors serialize as the raw bytes themselves — no length prefix, no delimiter.
 """
 
-from __future__ import annotations
-
 from collections.abc import Iterable
 from typing import IO, Any, ClassVar, Self, override
 
diff --git a/src/lean_spec/types/collections.py b/src/lean_spec/types/collections.py
index 26350bcf..2446a8b3 100644
--- a/src/lean_spec/types/collections.py
+++ b/src/lean_spec/types/collections.py
@@ -1,7 +1,5 @@
 """Vector and List Type Specifications."""
 
-from __future__ import annotations
-
 import io
 from collections.abc import Iterator, Sequence
 from typing import (
diff --git a/src/lean_spec/types/container.py b/src/lean_spec/types/container.py
index 9c6e9b1b..3f0718ea 100644
--- a/src/lean_spec/types/container.py
+++ b/src/lean_spec/types/container.py
@@ -7,8 +7,6 @@
 Ethereum's serialization format.
 """
 
-from __future__ import annotations
-
 from typing import IO, Any, Self, override
 
 from .constants import OFFSET_BYTE_LENGTH
diff --git a/src/lean_spec/types/ssz_base.py b/src/lean_spec/types/ssz_base.py
index 222e8b97..92b9fc90 100644
--- a/src/lean_spec/types/ssz_base.py
+++ b/src/lean_spec/types/ssz_base.py
@@ -1,7 +1,5 @@
 """Base classes and interfaces for all SSZ types."""
 
-from __future__ import annotations
-
 import io
 from abc import ABC, abstractmethod
 from collections.abc import Sequence