diff --git a/src/graphon/dsl/entities.py b/src/graphon/dsl/entities.py
index 3871b77..d9a17cc 100644
--- a/src/graphon/dsl/entities.py
+++ b/src/graphon/dsl/entities.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Mapping
 from enum import StrEnum, auto
 from typing import Any, Protocol
@@ -125,4 +126,5 @@ def loadable(self) -> bool:
 
 
 class TypedNodeFactory(Protocol):
+    @abc.abstractmethod
     def create_node(self, node_config: NodeConfigDict) -> Any: ...
diff --git a/src/graphon/dsl/slim/llm.py b/src/graphon/dsl/slim/llm.py
index 0bc8131..e31defb 100644
--- a/src/graphon/dsl/slim/llm.py
+++ b/src/graphon/dsl/slim/llm.py
@@ -3,7 +3,7 @@
 import json
 from collections.abc import Generator, Iterable, Mapping, Sequence
 from dataclasses import dataclass, field
-from typing import Any, Literal, overload, override
+from typing import TYPE_CHECKING, Any, Literal, overload, override
 
 from pydantic import StrictStr, TypeAdapter, ValidationError
 
@@ -682,3 +682,11 @@ def _parse_optional_llm_usage(payload: object) -> LLMUsage | None:
             raise TypeError(msg)
         normalized_payload[key] = value
     return LLMUsage.from_metadata(normalized_payload)
+
+
+if TYPE_CHECKING:
+    # static assertion to ensure SlimLLM implements LLMProtocol.
+    def _assert_slim_llm_protocol(
+        runtime: SlimLLM,
+    ) -> LLMProtocol:  # pyright: ignore[reportUnusedFunction]
+        return runtime
diff --git a/src/graphon/dsl/tool_runtime.py b/src/graphon/dsl/tool_runtime.py
index f389fbf..6623d5a 100644
--- a/src/graphon/dsl/tool_runtime.py
+++ b/src/graphon/dsl/tool_runtime.py
@@ -4,7 +4,7 @@
 from collections.abc import Callable, Generator, Iterable, Mapping, Sequence
 from dataclasses import dataclass, field
 from enum import StrEnum
-from typing import Any, cast
+from typing import TYPE_CHECKING, Any, cast
 
 from graphon.file.models import File
 from graphon.model_runtime.entities.llm_entities import LLMUsage
@@ -896,3 +896,12 @@ def _decode_blob(value: object) -> bytes:
             raise ToolNodeError(msg) from error
     msg = "Slim blob payload must be bytes or base64 text."
     raise ToolNodeError(msg)
+
+
+if TYPE_CHECKING:
+    # static assertion to ensure SlimToolNodeRuntime implements
+    # ToolNodeRuntimeProtocol.
+    def _assert_slim_tool_node_runtime_protocol(
+        runtime: SlimToolNodeRuntime,
+    ) -> ToolNodeRuntimeProtocol:  # pyright: ignore[reportUnusedFunction]
+        return runtime
diff --git a/src/graphon/file/protocols.py b/src/graphon/file/protocols.py
index c376c02..1caaecd 100644
--- a/src/graphon/file/protocols.py
+++ b/src/graphon/file/protocols.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Generator
 from typing import TYPE_CHECKING, Literal, Protocol
 
@@ -18,8 +19,10 @@ class WorkflowFileRuntimeProtocol(Protocol):
     """
 
     @property
+    @abc.abstractmethod
     def multimodal_send_format(self) -> str: ...
 
+    @abc.abstractmethod
     def http_get(
         self,
         url: str,
@@ -27,10 +30,13 @@ def http_get(
         follow_redirects: bool = True,
     ) -> HttpResponseProtocol: ...
 
+    @abc.abstractmethod
     def storage_load(self, path: str, *, stream: bool = False) -> bytes | Generator: ...
 
+    @abc.abstractmethod
     def load_file_bytes(self, *, file: File) -> bytes: ...
 
+    @abc.abstractmethod
     def resolve_file_url(
         self,
         *,
@@ -38,6 +44,7 @@ def resolve_file_url(
         for_external: bool = True,
     ) -> str | None: ...
 
+    @abc.abstractmethod
     def resolve_upload_file_url(
         self,
         *,
@@ -46,6 +53,7 @@ def resolve_upload_file_url(
         for_external: bool = True,
     ) -> str: ...
 
+    @abc.abstractmethod
     def resolve_tool_file_url(
         self,
         *,
@@ -54,6 +62,7 @@ def resolve_tool_file_url(
         for_external: bool = True,
     ) -> str: ...
 
+    @abc.abstractmethod
     def verify_preview_signature(
         self,
         *,
diff --git a/src/graphon/graph/graph.py b/src/graphon/graph/graph.py
index ac79390..2eb9481 100644
--- a/src/graphon/graph/graph.py
+++ b/src/graphon/graph/graph.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 import logging
 from collections import defaultdict
 from collections.abc import Mapping, Sequence
@@ -27,6 +28,7 @@ class NodeFactory(Protocol):
     allowing for different node creation strategies while maintaining type safety.
     """
 
+    @abc.abstractmethod
     def create_node(self, node_config: NodeConfigDict) -> Node:
         """Create a Node instance from node configuration data.
 
diff --git a/src/graphon/graph/validation.py b/src/graphon/graph/validation.py
index b66def8..af3ac0a 100644
--- a/src/graphon/graph/validation.py
+++ b/src/graphon/graph/validation.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Sequence
 from dataclasses import dataclass
 from typing import TYPE_CHECKING, Protocol
@@ -34,6 +35,7 @@ def __init__(self, issues: Sequence[GraphValidationIssue]) -> None:
 class GraphValidationRule(Protocol):
     """Protocol that individual validation rules must satisfy."""
 
+    @abc.abstractmethod
     def validate(self, graph: Graph) -> Sequence[GraphValidationIssue]:
         """Validate the provided graph and return any discovered issues."""
         ...
diff --git a/src/graphon/graph_engine/command_channels/in_memory_channel.py b/src/graphon/graph_engine/command_channels/in_memory_channel.py
index 557100e..ad91d00 100644
--- a/src/graphon/graph_engine/command_channels/in_memory_channel.py
+++ b/src/graphon/graph_engine/command_channels/in_memory_channel.py
@@ -5,7 +5,7 @@
 """
 
 from queue import Empty, Queue
-from typing import final
+from typing import TYPE_CHECKING, final
 
 from ..entities.commands import GraphEngineCommand
 
@@ -49,3 +49,13 @@ def send_command(self, command: GraphEngineCommand) -> None:
 
         """
         self._queue.put(command)
+
+
+if TYPE_CHECKING:
+    from .protocol import CommandChannel
+
+    # static assertion to ensure InMemoryChannel implements CommandChannel.
+    def _assert_command_channel(
+        channel: InMemoryChannel,
+    ) -> CommandChannel:  # pyright: ignore[reportUnusedFunction]
+        return channel
diff --git a/src/graphon/graph_engine/command_channels/protocol.py b/src/graphon/graph_engine/command_channels/protocol.py
index 10e3e59..956fd30 100644
--- a/src/graphon/graph_engine/command_channels/protocol.py
+++ b/src/graphon/graph_engine/command_channels/protocol.py
@@ -4,6 +4,7 @@
 to/from a GraphEngine instance, supporting both local and distributed scenarios.
 """
 
+import abc
 from typing import Protocol
 
 from ..entities.commands import GraphEngineCommand
@@ -16,6 +17,7 @@ class CommandChannel(Protocol):
     this channel is dedicated to that single execution.
     """
 
+    @abc.abstractmethod
     def fetch_commands(self) -> list[GraphEngineCommand]:
         """Fetch pending commands for this GraphEngine instance.
 
@@ -27,6 +29,7 @@ def fetch_commands(self) -> list[GraphEngineCommand]:
         """
         ...
 
+    @abc.abstractmethod
     def send_command(self, command: GraphEngineCommand) -> None:
         """Send a command to be processed by this GraphEngine instance.
 
diff --git a/src/graphon/graph_engine/command_channels/redis_channel.py b/src/graphon/graph_engine/command_channels/redis_channel.py
index 79312dc..8f45d27 100644
--- a/src/graphon/graph_engine/command_channels/redis_channel.py
+++ b/src/graphon/graph_engine/command_channels/redis_channel.py
@@ -5,9 +5,10 @@
 Each instance uses a unique key for its command queue.
 """
 
+import abc
 import json
 from contextlib import AbstractContextManager
-from typing import Any, Protocol, final
+from typing import TYPE_CHECKING, Any, Protocol, final
 
 from ..entities.commands import (
     AbortCommand,
@@ -27,18 +28,26 @@
 class RedisPipelineProtocol(Protocol):
     """Minimal Redis pipeline contract used by the command channel."""
 
+    @abc.abstractmethod
     def lrange(self, name: str, start: int, end: int) -> Any: ...
+    @abc.abstractmethod
     def delete(self, *names: str) -> Any: ...
+    @abc.abstractmethod
     def execute(self) -> list[Any]: ...
+    @abc.abstractmethod
     def rpush(self, name: str, *values: str) -> Any: ...
+    @abc.abstractmethod
     def expire(self, name: str, time: int) -> Any: ...
+    @abc.abstractmethod
     def set(self, name: str, value: str, ex: int | None = None) -> Any: ...
+    @abc.abstractmethod
     def get(self, name: str) -> Any: ...
 
 
 class RedisClientProtocol(Protocol):
     """Redis client contract required by the command channel."""
 
+    @abc.abstractmethod
     def pipeline(self) -> AbstractContextManager[RedisPipelineProtocol]: ...
 
 
@@ -160,3 +169,13 @@ def _has_pending_commands(self) -> bool:
             pending_value, _ = pipe.execute()
 
         return pending_value is not None
+
+
+if TYPE_CHECKING:
+    from .protocol import CommandChannel
+
+    # static assertion to ensure RedisChannel implements CommandChannel.
+    def _assert_command_channel(
+        channel: RedisChannel,
+    ) -> CommandChannel:  # pyright: ignore[reportUnusedFunction]
+        return channel
diff --git a/src/graphon/graph_engine/command_processing/command_handlers.py b/src/graphon/graph_engine/command_processing/command_handlers.py
index 745f8e4..978e25b 100644
--- a/src/graphon/graph_engine/command_processing/command_handlers.py
+++ b/src/graphon/graph_engine/command_processing/command_handlers.py
@@ -1,5 +1,5 @@
 import logging
-from typing import final, override
+from typing import TYPE_CHECKING, final, override
 
 from graphon.entities.pause_reason import SchedulingPause
 from graphon.runtime.graph_runtime_state import GraphExecutionProtocol
@@ -69,3 +69,21 @@ def handle(
                     execution.workflow_id,
                     exc,
                 )
+
+
+if TYPE_CHECKING:
+    # static assertions to ensure command handlers implement CommandHandler.
+    def _assert_abort_handler_protocol(
+        handler: AbortCommandHandler,
+    ) -> CommandHandler[AbortCommand]:  # pyright: ignore[reportUnusedFunction]
+        return handler
+
+    def _assert_pause_handler_protocol(
+        handler: PauseCommandHandler,
+    ) -> CommandHandler[PauseCommand]:  # pyright: ignore[reportUnusedFunction]
+        return handler
+
+    def _assert_update_variables_handler_protocol(
+        handler: UpdateVariablesCommandHandler,
+    ) -> CommandHandler[UpdateVariablesCommand]:  # pyright: ignore[reportUnusedFunction]
+        return handler
diff --git a/src/graphon/graph_engine/command_processing/command_processor.py b/src/graphon/graph_engine/command_processing/command_processor.py
index 99903d9..3be4317 100644
--- a/src/graphon/graph_engine/command_processing/command_processor.py
+++ b/src/graphon/graph_engine/command_processing/command_processor.py
@@ -1,5 +1,6 @@
 """Main command processor for handling external commands."""
 
+import abc
 import logging
 from collections.abc import Callable
 from typing import Protocol, final
@@ -15,6 +16,7 @@
 class CommandHandler[CommandT: GraphEngineCommand](Protocol):
     """Protocol for command handlers."""
 
+    @abc.abstractmethod
     def handle(
         self,
         command: CommandT,
diff --git a/src/graphon/graph_engine/domain/node_execution.py b/src/graphon/graph_engine/domain/node_execution.py
index 17270ee..aeb25c4 100644
--- a/src/graphon/graph_engine/domain/node_execution.py
+++ b/src/graphon/graph_engine/domain/node_execution.py
@@ -1,6 +1,7 @@
 """NodeExecution entity representing a node's execution state."""
 
 from dataclasses import dataclass
+from typing import TYPE_CHECKING
 
 from graphon.enums import NodeState
 
@@ -40,3 +41,13 @@ def mark_skipped(self) -> None:
     def increment_retry(self) -> None:
         """Increment the retry count for this node."""
         self.retry_count += 1
+
+
+if TYPE_CHECKING:
+    from graphon.runtime.graph_runtime_state import NodeExecutionProtocol
+
+    # static assertion to ensure NodeExecution implements NodeExecutionProtocol.
+    def _assert_node_execution_protocol(
+        execution: NodeExecution,
+    ) -> NodeExecutionProtocol:  # pyright: ignore[reportUnusedFunction]
+        return execution
diff --git a/src/graphon/graph_engine/ready_queue/in_memory.py b/src/graphon/graph_engine/ready_queue/in_memory.py
index 19dee5a..5677880 100644
--- a/src/graphon/graph_engine/ready_queue/in_memory.py
+++ b/src/graphon/graph_engine/ready_queue/in_memory.py
@@ -5,7 +5,7 @@
 """
 
 import queue
-from typing import final
+from typing import TYPE_CHECKING, final
 
 from .protocol import ReadyQueue, ReadyQueueState
 
@@ -137,3 +137,9 @@ def loads(self, data: str) -> None:
         # Restore items
         for item in state.items:
             self._queue.put(item)
+
+
+if TYPE_CHECKING:
+    # static assertion to ensure InMemoryReadyQueue implements ReadyQueue.
+    def _assert_ready_queue(queue_impl: InMemoryReadyQueue) -> ReadyQueue:  # pyright: ignore[reportUnusedFunction]
+        return queue_impl
diff --git a/src/graphon/graph_engine/ready_queue/protocol.py b/src/graphon/graph_engine/ready_queue/protocol.py
index 6c53677..cd1cb55 100644
--- a/src/graphon/graph_engine/ready_queue/protocol.py
+++ b/src/graphon/graph_engine/ready_queue/protocol.py
@@ -4,6 +4,7 @@
 for execution, supporting both in-memory and persistent storage scenarios.
 """
 
+import abc
 from collections.abc import Sequence
 from typing import Protocol
 
@@ -35,6 +36,7 @@ class ReadyQueue(Protocol):
     that can be serialized for state storage.
     """
 
+    @abc.abstractmethod
     def put(self, item: str) -> None:
         """Add a node ID to the ready queue.
 
@@ -44,6 +46,7 @@ def put(self, item: str) -> None:
         """
         ...
 
+    @abc.abstractmethod
     def get(self, timeout: float | None = None) -> str:
         """Retrieve and remove a node ID from the queue.
 
@@ -56,6 +59,7 @@ def get(self, timeout: float | None = None) -> str:
         """
         ...
 
+    @abc.abstractmethod
     def task_done(self) -> None:
         """Indicate that a previously retrieved task is complete.
 
@@ -64,6 +68,7 @@ def task_done(self) -> None:
         """
         ...
 
+    @abc.abstractmethod
     def empty(self) -> bool:
         """Check if the queue is empty.
 
@@ -73,6 +78,7 @@ def empty(self) -> bool:
         """
         ...
 
+    @abc.abstractmethod
     def qsize(self) -> int:
         """Get the approximate size of the queue.
 
@@ -82,6 +88,7 @@ def qsize(self) -> int:
         """
         ...
 
+    @abc.abstractmethod
     def dumps(self) -> str:
         """Serialize the queue state to a JSON string for storage.
 
@@ -92,6 +99,7 @@ def dumps(self) -> str:
         """
         ...
 
+    @abc.abstractmethod
     def loads(self, data: str) -> None:
         """Restore the queue state from a JSON string.
 
diff --git a/src/graphon/graph_engine/response_coordinator/coordinator.py b/src/graphon/graph_engine/response_coordinator/coordinator.py
index 0216b40..89a4e78 100644
--- a/src/graphon/graph_engine/response_coordinator/coordinator.py
+++ b/src/graphon/graph_engine/response_coordinator/coordinator.py
@@ -791,3 +791,14 @@ def loads(self, data: str) -> None:
                 if state.active_session
                 else None
             )
+
+
+if TYPE_CHECKING:
+    from graphon.runtime.graph_runtime_state import ResponseStreamCoordinatorProtocol
+
+    # static assertion to ensure ResponseStreamCoordinator implements
+    # ResponseStreamCoordinatorProtocol.
+    def _assert_response_stream_coordinator_protocol(
+        coordinator: ResponseStreamCoordinator,
+    ) -> ResponseStreamCoordinatorProtocol:  # pyright: ignore[reportUnusedFunction]
+        return coordinator
diff --git a/src/graphon/http/client.py b/src/graphon/http/client.py
index a258deb..514a397 100644
--- a/src/graphon/http/client.py
+++ b/src/graphon/http/client.py
@@ -1,7 +1,7 @@
 from __future__ import annotations
 
 from collections.abc import Sequence
-from typing import Any, override
+from typing import TYPE_CHECKING, Any, override
 
 import httpx
 
@@ -102,3 +102,11 @@ def _normalize_request_kwargs(kwargs: dict[str, Any]) -> dict[str, Any]:
             )
 
         return request_kwargs
+
+
+if TYPE_CHECKING:
+    # static assertion to ensure HttpxHttpClient implements HttpClientProtocol.
+    def _assert_http_client_protocol(
+        client: HttpxHttpClient,
+    ) -> HttpClientProtocol:  # pyright: ignore[reportUnusedFunction]
+        return client
diff --git a/src/graphon/http/protocols.py b/src/graphon/http/protocols.py
index 505dfa0..ce31c93 100644
--- a/src/graphon/http/protocols.py
+++ b/src/graphon/http/protocols.py
@@ -1,3 +1,4 @@
+import abc
 from collections.abc import Mapping
 from typing import Any, Protocol
 
@@ -6,38 +7,51 @@
 
 class HttpResponseProtocol(Protocol):
     @property
+    @abc.abstractmethod
     def headers(self) -> Mapping[str, str]: ...
 
     @property
+    @abc.abstractmethod
     def content(self) -> bytes: ...
 
     @property
+    @abc.abstractmethod
     def status_code(self) -> int: ...
 
     @property
+    @abc.abstractmethod
     def text(self) -> str: ...
 
     @property
+    @abc.abstractmethod
     def is_success(self) -> bool: ...
 
+    @abc.abstractmethod
     def raise_for_status(self) -> None: ...
 
 
 class HttpClientProtocol(Protocol):
     @property
+    @abc.abstractmethod
     def max_retries_exceeded_error(self) -> type[Exception]: ...
 
     @property
+    @abc.abstractmethod
     def request_error(self) -> type[Exception]: ...
 
+    @abc.abstractmethod
     def get(self, url: str, max_retries: int = ..., **kwargs: Any) -> HttpResponse: ...
 
+    @abc.abstractmethod
     def head(self, url: str, max_retries: int = ..., **kwargs: Any) -> HttpResponse: ...
 
+    @abc.abstractmethod
     def post(self, url: str, max_retries: int = ..., **kwargs: Any) -> HttpResponse: ...
 
+    @abc.abstractmethod
     def put(self, url: str, max_retries: int = ..., **kwargs: Any) -> HttpResponse: ...
 
+    @abc.abstractmethod
     def delete(
         self,
         url: str,
@@ -45,6 +59,7 @@ def delete(
         **kwargs: Any,
     ) -> HttpResponse: ...
 
+    @abc.abstractmethod
     def patch(
         self,
         url: str,
diff --git a/src/graphon/http/response.py b/src/graphon/http/response.py
index e033fd8..781ed5f 100644
--- a/src/graphon/http/response.py
+++ b/src/graphon/http/response.py
@@ -122,3 +122,13 @@ def _extract_charset_from_content_type(self) -> str | None:
     def raise_for_status(self) -> None:
         if not self.is_success:
             raise HttpStatusError(self)
+
+
+if TYPE_CHECKING:
+    from .protocols import HttpResponseProtocol
+
+    # static assertion to ensure HttpResponse implements HttpResponseProtocol.
+    def _assert_http_response_protocol(
+        response: HttpResponse,
+    ) -> HttpResponseProtocol:  # pyright: ignore[reportUnusedFunction]
+        return response
diff --git a/src/graphon/model_runtime/memory/prompt_message_memory.py b/src/graphon/model_runtime/memory/prompt_message_memory.py
index ea2c245..ccb6b36 100644
--- a/src/graphon/model_runtime/memory/prompt_message_memory.py
+++ b/src/graphon/model_runtime/memory/prompt_message_memory.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Sequence
 from typing import Protocol
 
@@ -11,6 +12,7 @@
 class PromptMessageMemory(Protocol):
     """Port for loading memory as prompt messages."""
 
+    @abc.abstractmethod
     def get_history_prompt_messages(
         self,
         max_token_limit: int = DEFAULT_MEMORY_MAX_TOKEN_LIMIT,
diff --git a/src/graphon/model_runtime/model_providers/base/tokenizers/gpt2_tokenizer.py b/src/graphon/model_runtime/model_providers/base/tokenizers/gpt2_tokenizer.py
index 2e33253..7bb3493 100644
--- a/src/graphon/model_runtime/model_providers/base/tokenizers/gpt2_tokenizer.py
+++ b/src/graphon/model_runtime/model_providers/base/tokenizers/gpt2_tokenizer.py
@@ -1,3 +1,4 @@
+import abc
 import logging
 from collections.abc import Sequence
 from pathlib import Path
@@ -8,6 +9,7 @@
 
 
 class _TokenizerProtocol(Protocol):
+    @abc.abstractmethod
     def encode(self, text: str) -> Sequence[int]: ...
 
 
diff --git a/src/graphon/model_runtime/protocols/llm_runtime.py b/src/graphon/model_runtime/protocols/llm_runtime.py
index 08b52d1..0e4c9ff 100644
--- a/src/graphon/model_runtime/protocols/llm_runtime.py
+++ b/src/graphon/model_runtime/protocols/llm_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Generator, Sequence
 from typing import Any, Literal, Protocol, overload, runtime_checkable
 
@@ -22,6 +23,7 @@ class LLMModelRuntime(ModelProviderRuntime, Protocol):
     """Runtime surface required by LLM-backed model wrappers."""
 
     @overload
+    @abc.abstractmethod
     def invoke_llm(
         self,
         *,
@@ -36,6 +38,7 @@ def invoke_llm(
     ) -> LLMResult: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm(
         self,
         *,
@@ -49,6 +52,7 @@ def invoke_llm(
         stream: Literal[True],
     ) -> Generator[LLMResultChunk, None, None]: ...
 
+    @abc.abstractmethod
     def invoke_llm(
         self,
         *,
@@ -63,6 +67,7 @@ def invoke_llm(
     ) -> LLMResult | Generator[LLMResultChunk, None, None]: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm_with_structured_output(
         self,
         *,
@@ -77,6 +82,7 @@ def invoke_llm_with_structured_output(
     ) -> LLMResultWithStructuredOutput: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm_with_structured_output(
         self,
         *,
@@ -90,6 +96,7 @@ def invoke_llm_with_structured_output(
         stream: Literal[True],
     ) -> Generator[LLMResultChunkWithStructuredOutput, None, None]: ...
 
+    @abc.abstractmethod
     def invoke_llm_with_structured_output(
         self,
         *,
@@ -106,6 +113,7 @@ def invoke_llm_with_structured_output(
         | Generator[LLMResultChunkWithStructuredOutput, None, None]
     ): ...
 
+    @abc.abstractmethod
     def get_llm_num_tokens(
         self,
         *,
diff --git a/src/graphon/model_runtime/protocols/moderation_runtime.py b/src/graphon/model_runtime/protocols/moderation_runtime.py
index 8cf6ae9..52c4087 100644
--- a/src/graphon/model_runtime/protocols/moderation_runtime.py
+++ b/src/graphon/model_runtime/protocols/moderation_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from typing import Any, Protocol, runtime_checkable
 
 from graphon.model_runtime.protocols.provider_runtime import ModelProviderRuntime
@@ -9,6 +10,7 @@
 class ModerationModelRuntime(ModelProviderRuntime, Protocol):
     """Runtime surface required by moderation model wrappers."""
 
+    @abc.abstractmethod
     def invoke_moderation(
         self,
         *,
diff --git a/src/graphon/model_runtime/protocols/provider_runtime.py b/src/graphon/model_runtime/protocols/provider_runtime.py
index 83c35fe..5cc98d7 100644
--- a/src/graphon/model_runtime/protocols/provider_runtime.py
+++ b/src/graphon/model_runtime/protocols/provider_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Sequence
 from typing import Any, Protocol, runtime_checkable
 
@@ -11,8 +12,10 @@
 class ModelProviderRuntime(Protocol):
     """Shared provider discovery, credential validation, and schema lookup."""
 
+    @abc.abstractmethod
     def fetch_model_providers(self) -> Sequence[ProviderEntity]: ...
 
+    @abc.abstractmethod
     def get_provider_icon(
         self,
         *,
@@ -21,6 +24,7 @@ def get_provider_icon(
         lang: str,
     ) -> tuple[bytes, str]: ...
 
+    @abc.abstractmethod
     def validate_provider_credentials(
         self,
         *,
@@ -28,6 +32,7 @@ def validate_provider_credentials(
         credentials: dict[str, Any],
     ) -> None: ...
 
+    @abc.abstractmethod
     def validate_model_credentials(
         self,
         *,
@@ -37,6 +42,7 @@ def validate_model_credentials(
         credentials: dict[str, Any],
     ) -> None: ...
 
+    @abc.abstractmethod
     def get_model_schema(
         self,
         *,
diff --git a/src/graphon/model_runtime/protocols/rerank_runtime.py b/src/graphon/model_runtime/protocols/rerank_runtime.py
index aa3814a..2f3a953 100644
--- a/src/graphon/model_runtime/protocols/rerank_runtime.py
+++ b/src/graphon/model_runtime/protocols/rerank_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from typing import Any, Protocol, runtime_checkable
 
 from graphon.model_runtime.entities.rerank_entities import (
@@ -13,6 +14,7 @@
 class RerankModelRuntime(ModelProviderRuntime, Protocol):
     """Runtime surface required by rerank model wrappers."""
 
+    @abc.abstractmethod
     def invoke_rerank(
         self,
         *,
@@ -25,6 +27,7 @@ def invoke_rerank(
         top_n: int | None,
     ) -> RerankResult: ...
 
+    @abc.abstractmethod
     def invoke_multimodal_rerank(
         self,
         *,
diff --git a/src/graphon/model_runtime/protocols/speech_to_text_runtime.py b/src/graphon/model_runtime/protocols/speech_to_text_runtime.py
index 8f59a62..07096a6 100644
--- a/src/graphon/model_runtime/protocols/speech_to_text_runtime.py
+++ b/src/graphon/model_runtime/protocols/speech_to_text_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from typing import IO, Any, Protocol, runtime_checkable
 
 from graphon.model_runtime.protocols.provider_runtime import ModelProviderRuntime
@@ -9,6 +10,7 @@
 class SpeechToTextModelRuntime(ModelProviderRuntime, Protocol):
     """Runtime surface required by speech-to-text model wrappers."""
 
+    @abc.abstractmethod
     def invoke_speech_to_text(
         self,
         *,
diff --git a/src/graphon/model_runtime/protocols/text_embedding_runtime.py b/src/graphon/model_runtime/protocols/text_embedding_runtime.py
index 4938ccd..b7808b8 100644
--- a/src/graphon/model_runtime/protocols/text_embedding_runtime.py
+++ b/src/graphon/model_runtime/protocols/text_embedding_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from typing import Any, Protocol, runtime_checkable
 
 from graphon.model_runtime.entities.text_embedding_entities import (
@@ -13,6 +14,7 @@
 class TextEmbeddingModelRuntime(ModelProviderRuntime, Protocol):
     """Runtime surface required by text and multimodal embedding wrappers."""
 
+    @abc.abstractmethod
     def invoke_text_embedding(
         self,
         *,
@@ -23,6 +25,7 @@ def invoke_text_embedding(
         input_type: EmbeddingInputType,
     ) -> EmbeddingResult: ...
 
+    @abc.abstractmethod
     def invoke_multimodal_embedding(
         self,
         *,
@@ -33,6 +36,7 @@ def invoke_multimodal_embedding(
         input_type: EmbeddingInputType,
     ) -> EmbeddingResult: ...
 
+    @abc.abstractmethod
     def get_text_embedding_num_tokens(
         self,
         *,
diff --git a/src/graphon/model_runtime/protocols/tts_runtime.py b/src/graphon/model_runtime/protocols/tts_runtime.py
index 2b9129a..37a924a 100644
--- a/src/graphon/model_runtime/protocols/tts_runtime.py
+++ b/src/graphon/model_runtime/protocols/tts_runtime.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Iterable
 from typing import Any, Protocol, runtime_checkable
 
@@ -10,6 +11,7 @@
 class TTSModelRuntime(ModelProviderRuntime, Protocol):
     """Runtime surface required by text-to-speech model wrappers."""
 
+    @abc.abstractmethod
     def invoke_tts(
         self,
         *,
@@ -20,6 +22,7 @@ def invoke_tts(
         voice: str,
     ) -> Iterable[bytes]: ...
 
+    @abc.abstractmethod
     def get_tts_model_voices(
         self,
         *,
diff --git a/src/graphon/nodes/code/code_node.py b/src/graphon/nodes/code/code_node.py
index a37039d..37c4c20 100644
--- a/src/graphon/nodes/code/code_node.py
+++ b/src/graphon/nodes/code/code_node.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Mapping, Sequence
 from decimal import Decimal
 from textwrap import dedent
@@ -23,6 +24,7 @@
 
 
 class CodeExecutorProtocol(Protocol):
+    @abc.abstractmethod
     def execute(
         self,
         *,
@@ -31,6 +33,7 @@ def execute(
         inputs: Mapping[str, Any],
     ) -> Mapping[str, Any]: ...
 
+    @abc.abstractmethod
     def is_execution_error(self, error: Exception) -> bool: ...
 
 
diff --git a/src/graphon/nodes/llm/file_saver.py b/src/graphon/nodes/llm/file_saver.py
index c6708ae..9372b1d 100644
--- a/src/graphon/nodes/llm/file_saver.py
+++ b/src/graphon/nodes/llm/file_saver.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 import mimetypes
 import typing as tp
 
@@ -21,6 +22,7 @@ class LLMFileSaver(tp.Protocol):
     LLM.
     """
 
+    @abc.abstractmethod
     def save_binary_string(
         self,
         data: bytes,
@@ -58,6 +60,7 @@ def save_binary_string(
         """
         raise NotImplementedError
 
+    @abc.abstractmethod
     def save_remote_url(self, url: str, file_type: FileType) -> File:
         """save_remote_url saves the file from a remote url returned by LLM.
 
@@ -198,3 +201,9 @@ def _validate_extension_override(extension_override: str | None) -> str | None:
             extension_override,
         )
     return extension_override
+
+
+if tp.TYPE_CHECKING:
+    # static assertion to ensure FileSaverImpl implements LLMFileSaver.
+    def _assert_llm_file_saver(saver: FileSaverImpl) -> LLMFileSaver:  # pyright: ignore[reportUnusedFunction]
+        return saver
diff --git a/src/graphon/nodes/llm/protocols.py b/src/graphon/nodes/llm/protocols.py
index 740d039..1c307f2 100644
--- a/src/graphon/nodes/llm/protocols.py
+++ b/src/graphon/nodes/llm/protocols.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from typing import Any, Protocol
 
 from graphon.nodes.llm.runtime_protocols import LLMProtocol
@@ -8,6 +9,7 @@
 class CredentialsProvider(Protocol):
     """Port for loading runtime credentials for a provider/model pair."""
 
+    @abc.abstractmethod
     def fetch(self, provider_name: str, model_name: str) -> dict[str, Any]:
         """Return credentials for the target provider/model or raise a domain error."""
         ...
@@ -16,6 +18,7 @@ def fetch(self, provider_name: str, model_name: str) -> dict[str, Any]:
 class ModelFactory(Protocol):
     """Port for creating prepared graph-facing LLM runtimes for execution."""
 
+    @abc.abstractmethod
     def init_model_instance(
         self,
         provider_name: str,
diff --git a/src/graphon/nodes/llm/runtime_protocols.py b/src/graphon/nodes/llm/runtime_protocols.py
index efafc4c..3b31cb3 100644
--- a/src/graphon/nodes/llm/runtime_protocols.py
+++ b/src/graphon/nodes/llm/runtime_protocols.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 from collections.abc import Generator, Mapping, Sequence
 from typing import Any, Literal, Protocol, overload
 
@@ -22,25 +23,33 @@ class LLMProtocol(Protocol):
     """A graph-facing LLM runtime adapter for node execution."""
 
     @property
+    @abc.abstractmethod
     def provider(self) -> str: ...
 
     @property
+    @abc.abstractmethod
     def model_name(self) -> str: ...
 
     @property
+    @abc.abstractmethod
     def parameters(self) -> Mapping[str, Any]: ...
 
     @parameters.setter
+    @abc.abstractmethod
     def parameters(self, value: Mapping[str, Any]) -> None: ...
 
     @property
+    @abc.abstractmethod
     def stop(self) -> Sequence[str] | None: ...
 
+    @abc.abstractmethod
     def get_model_schema(self) -> AIModelEntity: ...
 
+    @abc.abstractmethod
     def get_llm_num_tokens(self, prompt_messages: Sequence[PromptMessage]) -> int: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm(
         self,
         *,
@@ -52,6 +61,7 @@ def invoke_llm(
     ) -> LLMResult: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm(
         self,
         *,
@@ -62,6 +72,7 @@ def invoke_llm(
         stream: Literal[True],
     ) -> Generator[LLMResultChunk, None, None]: ...
 
+    @abc.abstractmethod
     def invoke_llm(
         self,
         *,
@@ -73,6 +84,7 @@ def invoke_llm(
     ) -> LLMResult | Generator[LLMResultChunk, None, None]: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm_with_structured_output(
         self,
         *,
@@ -84,6 +96,7 @@ def invoke_llm_with_structured_output(
     ) -> LLMResultWithStructuredOutput: ...
 
     @overload
+    @abc.abstractmethod
     def invoke_llm_with_structured_output(
         self,
         *,
@@ -94,6 +107,7 @@ def invoke_llm_with_structured_output(
         stream: Literal[True],
     ) -> Generator[LLMResultChunkWithStructuredOutput, None, None]: ...
 
+    @abc.abstractmethod
     def invoke_llm_with_structured_output(
         self,
         *,
@@ -107,12 +121,14 @@ def invoke_llm_with_structured_output(
         | Generator[LLMResultChunkWithStructuredOutput, None, None]
     ): ...
 
+    @abc.abstractmethod
     def is_structured_output_parse_error(self, error: Exception) -> bool: ...
 
 
 class PromptMessageSerializerProtocol(Protocol):
     """Port for converting compiled prompt messages into persisted process data."""
 
+    @abc.abstractmethod
     def serialize(
         self,
         *,
@@ -124,4 +140,5 @@ def serialize(
 class RetrieverAttachmentLoaderProtocol(Protocol):
     """Port for resolving retriever segment attachments into graph file references."""
 
+    @abc.abstractmethod
     def load(self, *, segment_id: str) -> Sequence[File]: ...
diff --git a/src/graphon/nodes/protocols.py b/src/graphon/nodes/protocols.py
index 3557792..df903d6 100644
--- a/src/graphon/nodes/protocols.py
+++ b/src/graphon/nodes/protocols.py
@@ -1,3 +1,4 @@
+import abc
 from collections.abc import Generator, Mapping
 from typing import Any, Protocol
 
@@ -6,10 +7,12 @@
 
 
 class FileManagerProtocol(Protocol):
+    @abc.abstractmethod
     def download(self, f: File, /) -> bytes: ...
 
 
 class ToolFileManagerProtocol(Protocol):
+    @abc.abstractmethod
     def create_file_by_raw(
         self,
         *,
@@ -18,6 +21,7 @@ def create_file_by_raw(
         filename: str | None = None,
     ) -> Any: ...
 
+    @abc.abstractmethod
     def get_file_generator_by_tool_file_id(
         self,
         tool_file_id: str,
@@ -29,6 +33,7 @@ class FileReferenceFactoryProtocol(Protocol):
     format. It enforces approriate permission filtering for the file.
     """
 
+    @abc.abstractmethod
     def build_from_mapping(self, *, mapping: Mapping[str, Any]) -> File: ...
 
 
diff --git a/src/graphon/nodes/runtime.py b/src/graphon/nodes/runtime.py
index 2843e4a..fb35f66 100644
--- a/src/graphon/nodes/runtime.py
+++ b/src/graphon/nodes/runtime.py
@@ -26,6 +26,7 @@ class ToolNodeRuntimeProtocol(Protocol):
     translate between graph-owned abstractions and `core.tools` internals.
     """
 
+    @abc.abstractmethod
     def get_runtime(
         self,
         *,
@@ -35,12 +36,14 @@ def get_runtime(
         node_execution_id: str | None = None,
     ) -> ToolRuntimeHandle: ...
 
+    @abc.abstractmethod
     def get_runtime_parameters(
         self,
         *,
         tool_runtime: ToolRuntimeHandle,
     ) -> Sequence[ToolRuntimeParameter]: ...
 
+    @abc.abstractmethod
     def invoke(
         self,
         *,
@@ -50,12 +53,14 @@ def invoke(
         provider_name: str,
     ) -> Generator[ToolRuntimeMessage, None, None]: ...
 
+    @abc.abstractmethod
     def get_usage(
         self,
         *,
         tool_runtime: ToolRuntimeHandle,
     ) -> LLMUsage: ...
 
+    @abc.abstractmethod
     def build_file_reference(self, *, mapping: Mapping[str, Any]) -> Any: ...
 
 
@@ -85,6 +90,7 @@ def create_form(
 class HumanInputFormRepositoryBindableRuntimeProtocol(Protocol):
     """Optional capability for runtimes that require explicit repository binding."""
 
+    @abc.abstractmethod
     def with_form_repository(
         self,
         form_repository: object,
@@ -130,22 +136,29 @@ def _normalize_human_input_runtime(
 
 class HumanInputFormStateProtocol(Protocol):
     @property
+    @abc.abstractmethod
     def id(self) -> str: ...
 
     @property
+    @abc.abstractmethod
     def rendered_content(self) -> str: ...
 
     @property
+    @abc.abstractmethod
     def selected_action_id(self) -> str | None: ...
 
     @property
+    @abc.abstractmethod
     def submitted_data(self) -> Mapping[str, Any] | None: ...
 
     @property
+    @abc.abstractmethod
     def submitted(self) -> bool: ...
 
     @property
+    @abc.abstractmethod
     def status(self) -> HumanInputFormStatus: ...
 
     @property
+    @abc.abstractmethod
     def expiration_time(self) -> datetime: ...
diff --git a/src/graphon/runtime/graph_runtime_state.py b/src/graphon/runtime/graph_runtime_state.py
index c4a9221..b29f57b 100644
--- a/src/graphon/runtime/graph_runtime_state.py
+++ b/src/graphon/runtime/graph_runtime_state.py
@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import abc
 import importlib
 import json
 from collections.abc import Mapping, Sequence
@@ -25,32 +26,39 @@
 class ReadyQueueProtocol(Protocol):
     """Structural interface required from ready queue implementations."""
 
+    @abc.abstractmethod
     def put(self, item: str) -> None:
         """Enqueue the identifier of a node that is ready to run."""
         ...
 
+    @abc.abstractmethod
     def get(self, timeout: float | None = None) -> str:
         """Return the next node identifier,
         blocking until available or timeout expires.
         """
         ...
 
+    @abc.abstractmethod
     def task_done(self) -> None:
         """Signal that the most recently dequeued node has completed processing."""
         ...
 
+    @abc.abstractmethod
     def empty(self) -> bool:
         """Return True when the queue contains no pending nodes."""
         ...
 
+    @abc.abstractmethod
     def qsize(self) -> int:
         """Approximate the number of pending nodes awaiting execution."""
         ...
 
+    @abc.abstractmethod
     def dumps(self) -> str:
         """Serialize the queue contents for persistence."""
         ...
 
+    @abc.abstractmethod
     def loads(self, data: str) -> None:
         """Restore the queue contents from a serialized payload."""
         ...
@@ -63,18 +71,22 @@ class NodeExecutionProtocol(Protocol):
     retry_count: int
     execution_id: str | None
 
+    @abc.abstractmethod
     def mark_started(self, execution_id: str) -> None:
         """Mark the node execution as started."""
         ...
 
+    @abc.abstractmethod
     def mark_taken(self) -> None:
         """Mark the node execution as successfully completed."""
         ...
 
+    @abc.abstractmethod
     def mark_failed(self, error: str) -> None:
         """Mark the node execution as failed with an error."""
         ...
 
+    @abc.abstractmethod
     def increment_retry(self) -> None:
         """Increment the retry counter for the node execution."""
         ...
@@ -98,52 +110,64 @@ class GraphExecutionProtocol(Protocol):
     pause_reasons: list[PauseReason]
 
     @property
+    @abc.abstractmethod
     def node_executions(self) -> Mapping[str, NodeExecutionProtocol]:
         """Return the persisted node execution state keyed by node id."""
         ...
 
+    @abc.abstractmethod
     def start(self) -> None:
         """Transition execution into the running state."""
         ...
 
+    @abc.abstractmethod
     def complete(self) -> None:
         """Mark execution as successfully completed."""
         ...
 
+    @abc.abstractmethod
     def abort(self, reason: str) -> None:
         """Abort execution in response to an external stop request."""
         ...
 
+    @abc.abstractmethod
     def pause(self, reason: PauseReason) -> None:
         """Pause execution with a recorded reason."""
         ...
 
+    @abc.abstractmethod
     def fail(self, error: Exception) -> None:
         """Record an unrecoverable error and end execution."""
         ...
 
+    @abc.abstractmethod
     def record_node_failure(self) -> None:
         """Increment the count of node failures observed during execution."""
         ...
 
+    @abc.abstractmethod
     def get_or_create_node_execution(self, node_id: str) -> NodeExecutionProtocol:
         """Return the execution entity for a node, creating it when needed."""
         ...
 
     @property
+    @abc.abstractmethod
     def is_paused(self) -> bool:
         """Return whether the execution is currently paused."""
         ...
 
     @property
+    @abc.abstractmethod
     def has_error(self) -> bool:
         """Return whether the execution has recorded an error."""
         ...
 
+    @abc.abstractmethod
     def dumps(self) -> str:
         """Serialize execution state into a JSON payload."""
         ...
 
+    @abc.abstractmethod
     def loads(self, data: str) -> None:
         """Restore execution state from a previously serialized payload."""
         ...
@@ -152,18 +176,22 @@ def loads(self, data: str) -> None:
 class ResponseStreamCoordinatorProtocol(Protocol):
     """Structural interface for response stream coordinator."""
 
+    @abc.abstractmethod
     def register(self, response_node_id: str) -> None:
         """Register a response node so its outputs can be streamed."""
         ...
 
+    @abc.abstractmethod
     def track_node_execution(self, node_id: str, execution_id: str) -> None:
         """Track the current execution id for a node."""
         ...
 
+    @abc.abstractmethod
     def on_edge_taken(self, edge_id: str) -> Sequence[NodeRunStreamChunkEvent]:
         """Update pending response sessions after an edge is taken."""
         ...
 
+    @abc.abstractmethod
     def intercept_event(
         self,
         event: NodeRunStreamChunkEvent | NodeRunSucceededEvent,
@@ -171,10 +199,12 @@ def intercept_event(
         """Translate node events into streamed response events."""
         ...
 
+    @abc.abstractmethod
     def loads(self, data: str) -> None:
         """Restore coordinator state from a serialized payload."""
         ...
 
+    @abc.abstractmethod
     def dumps(self) -> str:
         """Serialize coordinator state for persistence."""
         ...
@@ -188,6 +218,7 @@ class NodeProtocol(Protocol):
     execution_type: NodeExecutionType
     node_type: ClassVar[NodeType]
 
+    @abc.abstractmethod
     def blocks_variable_output(
         self,
         variable_selectors: set[tuple[str, ...]],
@@ -211,10 +242,12 @@ class GraphProtocol(Protocol):
     edges: Mapping[str, EdgeProtocol]
     root_node: NodeProtocol
 
+    @abc.abstractmethod
     def get_outgoing_edges(self, node_id: str) -> Sequence[EdgeProtocol]: ...
 
 
 class ChildGraphEngineBuilderProtocol(Protocol):
+    @abc.abstractmethod
     def build_child_engine(
         self,
         *,
diff --git a/src/graphon/runtime/graph_runtime_state_protocol.py b/src/graphon/runtime/graph_runtime_state_protocol.py
index 7b5c9ce..e732e78 100644
--- a/src/graphon/runtime/graph_runtime_state_protocol.py
+++ b/src/graphon/runtime/graph_runtime_state_protocol.py
@@ -1,3 +1,4 @@
+import abc
 from collections.abc import Mapping, Sequence
 from typing import Protocol
 
@@ -8,10 +9,12 @@
 class ReadOnlyVariablePool(Protocol):
     """Read-only interface for VariablePool."""
 
+    @abc.abstractmethod
     def get(self, selector: Sequence[str], /) -> Segment | None:
         """Get a variable value (read-only)."""
         ...
 
+    @abc.abstractmethod
     def get_by_prefix(self, prefix: str, /) -> Mapping[str, object]:
         """Get all variables stored under a given node prefix (read-only)."""
         ...
@@ -26,49 +29,59 @@ class ReadOnlyGraphRuntimeState(Protocol):
     """
 
     @property
+    @abc.abstractmethod
     def variable_pool(self) -> ReadOnlyVariablePool:
         """Get read-only access to the variable pool."""
         ...
 
     @property
+    @abc.abstractmethod
     def start_at(self) -> float:
         """Get the start time (read-only)."""
         ...
 
     @property
+    @abc.abstractmethod
     def total_tokens(self) -> int:
         """Get the total tokens count (read-only)."""
         ...
 
     @property
+    @abc.abstractmethod
     def llm_usage(self) -> LLMUsage:
         """Get a copy of LLM usage info (read-only)."""
         ...
 
     @property
+    @abc.abstractmethod
     def outputs(self) -> dict[str, object]:
         """Get a defensive copy of outputs (read-only)."""
         ...
 
     @property
+    @abc.abstractmethod
     def node_run_steps(self) -> int:
         """Get the node run steps count (read-only)."""
         ...
 
     @property
+    @abc.abstractmethod
     def ready_queue_size(self) -> int:
         """Get the number of nodes currently in the ready queue."""
         ...
 
     @property
+    @abc.abstractmethod
     def exceptions_count(self) -> int:
         """Get the number of node execution exceptions recorded."""
         ...
 
+    @abc.abstractmethod
     def get_output(self, key: str, default: object = None) -> object:
         """Get a single output value (returns a copy)."""
         ...
 
+    @abc.abstractmethod
     def dumps(self) -> str:
         """Serialize the runtime state into a JSON snapshot (read-only)."""
         ...
diff --git a/src/graphon/runtime/read_only_wrappers.py b/src/graphon/runtime/read_only_wrappers.py
index 822c3ce..8c836fa 100644
--- a/src/graphon/runtime/read_only_wrappers.py
+++ b/src/graphon/runtime/read_only_wrappers.py
@@ -2,6 +2,7 @@
 
 from collections.abc import Mapping, Sequence
 from copy import deepcopy
+from typing import TYPE_CHECKING
 
 from graphon.model_runtime.entities.llm_entities import LLMUsage
 from graphon.variables.segments import Segment
@@ -72,3 +73,18 @@ def get_output(self, key: str, default: object = None) -> object:
     def dumps(self) -> str:
         """Serialize the underlying runtime state for external persistence."""
         return self._state.dumps()
+
+
+if TYPE_CHECKING:
+    from .graph_runtime_state_protocol import ReadOnlyGraphRuntimeState
+
+    # static assertions to ensure read-only wrappers implement their protocols.
+    def _assert_readonly_variable_pool_wrapper(
+        pool: ReadOnlyVariablePoolWrapper,
+    ) -> ReadOnlyVariablePool:  # pyright: ignore[reportUnusedFunction]
+        return pool
+
+    def _assert_readonly_graph_runtime_state_wrapper(
+        state: ReadOnlyGraphRuntimeStateWrapper,
+    ) -> ReadOnlyGraphRuntimeState:  # pyright: ignore[reportUnusedFunction]
+        return state
diff --git a/src/graphon/variable_loader.py b/src/graphon/variable_loader.py
index 3cf32ca..3c97633 100644
--- a/src/graphon/variable_loader.py
+++ b/src/graphon/variable_loader.py
@@ -1,6 +1,6 @@
 import abc
 from collections.abc import Mapping, Sequence
-from typing import Any, Protocol
+from typing import TYPE_CHECKING, Any, Protocol
 
 from graphon.runtime.variable_pool import VariablePool
 from graphon.variables.consts import SELECTORS_LENGTH
@@ -45,6 +45,14 @@ def load_variables(self, selectors: list[list[str]]) -> list[VariableBase]:
         return []
 
 
+if TYPE_CHECKING:
+    # static assertion to ensure _DummyVariableLoader implements VariableLoader.
+    def _assert_variable_loader(
+        loader: _DummyVariableLoader,
+    ) -> VariableLoader:  # pyright: ignore[reportUnusedFunction]
+        return loader
+
+
 DUMMY_VARIABLE_LOADER = _DummyVariableLoader()
 
 
diff --git a/tests/test_protocol_abstract_contracts.py b/tests/test_protocol_abstract_contracts.py
new file mode 100644
index 0000000..0468b33
--- /dev/null
+++ b/tests/test_protocol_abstract_contracts.py
@@ -0,0 +1,184 @@
+from __future__ import annotations
+
+import ast
+import importlib
+import inspect
+from pathlib import Path
+from typing import Any
+
+import pytest
+
+
+def _is_direct_protocol_class(
+    class_def: ast.ClassDef,
+    *,
+    protocol_aliases: set[str],
+) -> bool:
+    for base in class_def.bases:
+        if isinstance(base, ast.Name) and base.id in protocol_aliases:
+            return True
+        if (
+            isinstance(base, ast.Attribute)
+            and isinstance(base.value, ast.Name)
+            and f"{base.value.id}.{base.attr}" in protocol_aliases
+        ):
+            return True
+    return False
+
+
+def _discover_protocol_aliases(parsed: ast.Module) -> set[str]:
+    protocol_aliases = {"Protocol"}
+    typing_aliases: set[str] = set()
+    for node in parsed.body:
+        if isinstance(node, ast.ImportFrom) and node.module == "typing":
+            for alias in node.names:
+                if alias.name == "Protocol":
+                    protocol_aliases.add(alias.asname or alias.name)
+        elif isinstance(node, ast.Import):
+            for alias in node.names:
+                if alias.name == "typing":
+                    typing_aliases.add(alias.asname or "typing")
+
+    protocol_aliases.update(f"{alias}.Protocol" for alias in typing_aliases)
+    return protocol_aliases
+
+
+def _has_protocol_members(class_def: ast.ClassDef) -> bool:
+    return any(
+        isinstance(member, ast.FunctionDef | ast.AsyncFunctionDef)
+        for member in class_def.body
+    )
+
+
+def _discover_protocol_targets() -> list[type[object]]:
+    src_root = Path(__file__).resolve().parents[1] / "src" / "graphon"
+    protocol_classes: list[type[object]] = []
+
+    for file_path in sorted(src_root.rglob("*.py")):
+        parsed = ast.parse(file_path.read_text())
+        protocol_aliases = _discover_protocol_aliases(parsed)
+        module_name = "graphon." + ".".join(
+            file_path.relative_to(src_root).with_suffix("").parts
+        )
+        module = importlib.import_module(module_name)
+
+        for class_def in [
+            node for node in parsed.body if isinstance(node, ast.ClassDef)
+        ]:
+            if not _is_direct_protocol_class(
+                class_def,
+                protocol_aliases=protocol_aliases,
+            ):
+                continue
+            if not _has_protocol_members(class_def):
+                continue
+            protocol_classes.append(getattr(module, class_def.name))
+
+    protocol_classes.sort(key=lambda cls: (cls.__module__, cls.__name__))
+    return protocol_classes
+
+
+def _protocol_member_names(protocol_cls: type[object]) -> list[str]:
+    member_names: list[str] = []
+    for name, value in protocol_cls.__dict__.items():
+        if name.startswith("__") and name.endswith("__"):
+            continue
+        if isinstance(value, property | classmethod | staticmethod):
+            member_names.append(name)
+            continue
+        if inspect.isfunction(value):
+            member_names.append(name)
+    return member_names
+
+
+def _build_member_override(protocol_cls: type[object], member_name: str) -> Any:
+    member = protocol_cls.__dict__[member_name]
+    if isinstance(member, property):
+        return property(lambda _: None)
+    if isinstance(member, classmethod):
+
+        def _class_stub(_cls: type[object], *args: object, **kwargs: object) -> None:
+            _ = args, kwargs
+
+        return classmethod(_class_stub)
+    if isinstance(member, staticmethod):
+
+        def _static_stub(*args: object, **kwargs: object) -> None:
+            _ = args, kwargs
+
+        return staticmethod(_static_stub)
+
+    def _stub(self: object, *args: object, **kwargs: object) -> None:
+        _ = self, args, kwargs
+
+    return _stub
+
+
+PROTOCOL_TARGETS = _discover_protocol_targets()
+
+
+@pytest.mark.parametrize(
+    "protocol_cls",
+    PROTOCOL_TARGETS,
+    ids=lambda cls: f"{cls.__module__}.{cls.__name__}",
+)
+def test_protocol_members_are_abstract(protocol_cls: type[object]) -> None:
+    member_names = _protocol_member_names(protocol_cls)
+    assert member_names, (
+        f"{protocol_cls.__module__}.{protocol_cls.__name__} has no protocol members."
+    )
+
+    non_abstract_members = [
+        name
+        for name in member_names
+        if not getattr(protocol_cls.__dict__[name], "__isabstractmethod__", False)
+    ]
+    assert not non_abstract_members, (
+        f"{protocol_cls.__module__}.{protocol_cls.__name__} contains non-abstract "
+        f"members: {non_abstract_members!r}"
+    )
+
+
+@pytest.mark.parametrize(
+    "protocol_cls",
+    PROTOCOL_TARGETS,
+    ids=lambda cls: f"{cls.__module__}.{cls.__name__}",
+)
+def test_protocol_direct_subclass_requires_overrides(
+    protocol_cls: type[object],
+) -> None:
+    direct_impl = type(
+        f"Direct{protocol_cls.__name__}",
+        (protocol_cls,),
+        {},
+    )
+    with pytest.raises(TypeError):
+        direct_impl()
+
+
+@pytest.mark.parametrize(
+    "protocol_cls",
+    PROTOCOL_TARGETS,
+    ids=lambda cls: f"{cls.__module__}.{cls.__name__}",
+)
+def test_protocol_indirect_partial_override_stays_abstract(
+    protocol_cls: type[object],
+) -> None:
+    member_names = _protocol_member_names(protocol_cls)
+    if len(member_names) < 2:
+        pytest.skip("Protocol has fewer than two members.")
+
+    first_member = member_names[0]
+    intermediate = type(
+        f"Intermediate{protocol_cls.__name__}",
+        (protocol_cls,),
+        {},
+    )
+    partial_impl = type(
+        f"Partial{protocol_cls.__name__}",
+        (intermediate,),
+        {first_member: _build_member_override(protocol_cls, first_member)},
+    )
+
+    with pytest.raises(TypeError):
+        partial_impl()