refactor(api): type LLM generator results with TypedDict

[statxc]

Part of #32863 (api/core/llm_generator/llm_generator.py)

Summary

• Add CodeGenerateResultDict TypedDict with keys code, language, error
• Add StructuredOutputResultDict TypedDict with keys output, error
• Annotate return types of LLMGenerator.generate_code and LLMGenerator.generate_structured_output

Why this change

generate_code and generate_structured_output return dicts with well-known fixed keys but had no return type annotations. The TypedDicts make the shape explicit.

Changes

• api/core/llm_generator/llm_generator.py: Define 2 TypedDicts, update 2 method signatures

[github-actions]

Pyrefly Diff

base → PR

--- /tmp/pyrefly_base.txt	2026-04-07 00:13:02.225351788 +0000
+++ /tmp/pyrefly_pr.txt	2026-04-07 00:12:53.746314899 +0000
@@ -23,23 +23,23 @@
 ERROR `defaultdict[str, dict[Unknown, Unknown]]` is not assignable to `defaultdict[ModelType, dict[str, ModelSettings]]` [bad-assignment]
     --> core/entities/provider_configuration.py:1531:79
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, int], stream=Literal[False]) [no-matching-overload]
-  --> core/llm_generator/llm_generator.py:74:60
+  --> core/llm_generator/llm_generator.py:85:60
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, float | int], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:137:60
+   --> core/llm_generator/llm_generator.py:148:60
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, Any], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:186:64
+   --> core/llm_generator/llm_generator.py:197:64
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, Any], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:234:70
+   --> core/llm_generator/llm_generator.py:245:70
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, Any], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:265:73
+   --> core/llm_generator/llm_generator.py:276:73
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, Any], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:274:73
+   --> core/llm_generator/llm_generator.py:285:73
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[UserPromptMessage], model_parameters=dict[str, Any], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:321:60
+   --> core/llm_generator/llm_generator.py:332:60
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[SystemPromptMessage | UserPromptMessage], model_parameters=dict[str, Any], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:381:60
+   --> core/llm_generator/llm_generator.py:394:60
 ERROR No matching overload found for function `core.model_manager.ModelInstance.invoke_llm` called with arguments: (prompt_messages=list[SystemPromptMessage | UserPromptMessage], model_parameters=dict[str, float], stream=Literal[False]) [no-matching-overload]
-   --> core/llm_generator/llm_generator.py:569:60
+   --> core/llm_generator/llm_generator.py:582:60
 ERROR Argument `dict[str, list[str] | str | None]` is not assignable to parameter `attributes` with type `dict[str, str] | None` in function `mlflow.tracing.fluent.start_span_no_context` [bad-argument-type]
    --> core/ops/mlflow_trace/mlflow_trace.py:271:24
 ERROR Argument `dict[str, dict[str, Any] | str | None]` is not assignable to parameter `attributes` with type `dict[str, str] | None` in function `mlflow.tracing.fluent.start_span_no_context` [bad-argument-type]

[statxc]

Hi @asukaminato0721 How are you? Could you help me with passing CI test failure?

[statxc]

Thank you @asukaminato0721