feat: add GPT-5-Codex support with Responses API integration

Adds support for OpenAI's GPT-5-Codex model which uses the new Responses API
endpoint (/v1/responses) instead of the standard Chat Completions API.

Changes:
- Add GPT-5-Codex to MODEL_CAPABILITIES with 400K context, 128K output
- Prioritize GPT-5-Codex for EXTENDED_REASONING tasks
- Add aliases: codex, gpt5-codex, gpt-5-code
- Update tests to expect GPT-5-Codex for extended reasoning

Benefits:
- 40-80% cost savings through Responses API caching
- 3% better performance on coding tasks (SWE-bench)
- Leverages existing dual-API infrastructure

providers/openai_provider.py

@@ -174,6 +174,25 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
             description="GPT-4.1 (1M context) - Advanced reasoning model with large context window",
             aliases=["gpt4.1"],
         ),
+        "gpt-5-codex": ModelCapabilities(
+            provider=ProviderType.OPENAI,
+            model_name="gpt-5-codex",
+            friendly_name="OpenAI (GPT-5 Codex)",
+            intelligence_score=17,  # Higher than GPT-5 for coding tasks
+            context_window=400_000,  # 400K tokens (same as GPT-5)
+            max_output_tokens=128_000,  # 128K output tokens
+            supports_extended_thinking=True,  # Responses API supports reasoning tokens
+            supports_system_prompts=True,
+            supports_streaming=True,
+            supports_function_calling=True,  # Enhanced for agentic software engineering
+            supports_json_mode=True,
+            supports_images=True,  # Screenshots, wireframes, diagrams
+            max_image_size_mb=20.0,  # 20MB per OpenAI docs
+            supports_temperature=True,
+            temperature_constraint=TemperatureConstraint.create("range"),
+            description="GPT-5 Codex (400K context) - Uses Responses API for 40-80% cost savings. Specialized for coding, refactoring, and software architecture. 3% better performance on SWE-bench.",
+            aliases=["gpt5-codex", "codex", "gpt-5-code", "gpt5-code"],
+        ),
     }
 
     def __init__(self, api_key: str, **kwargs):
@@ -290,15 +309,18 @@ class OpenAIModelProvider(OpenAICompatibleProvider):
 
         if category == ToolModelCategory.EXTENDED_REASONING:
             # Prefer models with extended thinking support
-            preferred = find_first(["o3", "o3-pro", "gpt-5"])
+            # GPT-5-Codex first for coding tasks (uses Responses API with 40-80% cost savings)
+            preferred = find_first(["gpt-5-codex", "o3", "o3-pro", "gpt-5"])
             return preferred if preferred else allowed_models[0]
 
         elif category == ToolModelCategory.FAST_RESPONSE:
             # Prefer fast, cost-efficient models
-            preferred = find_first(["gpt-5", "gpt-5-mini", "o4-mini", "o3-mini"])
+            # GPT-5 models for speed, GPT-5-Codex after (premium pricing but cached)
+            preferred = find_first(["gpt-5", "gpt-5-mini", "gpt-5-codex", "o4-mini", "o3-mini"])
             return preferred if preferred else allowed_models[0]
 
         else:  # BALANCED or default
             # Prefer balanced performance/cost models
-            preferred = find_first(["gpt-5", "gpt-5-mini", "o4-mini", "o3-mini"])
+            # Include GPT-5-Codex for coding workflows
+            preferred = find_first(["gpt-5", "gpt-5-codex", "gpt-5-mini", "o4-mini", "o3-mini"])
             return preferred if preferred else allowed_models[0]

tests/test_auto_mode_comprehensive.py

@@ -93,7 +93,7 @@ class TestAutoModeComprehensive:
                     "OPENROUTER_API_KEY": None,
                 },
                 {
-                    "EXTENDED_REASONING": "o3",  # O3 for deep reasoning
+                    "EXTENDED_REASONING": "gpt-5-codex",  # GPT-5-Codex prioritized for coding tasks
                     "FAST_RESPONSE": "gpt-5",  # Prefer gpt-5 for speed
                     "BALANCED": "gpt-5",  # Prefer gpt-5 for balanced
                 },

tests/test_auto_mode_provider_selection.py

@@ -98,7 +98,7 @@ class TestAutoModeProviderSelection:
             balanced = ModelProviderRegistry.get_preferred_fallback_model(ToolModelCategory.BALANCED)
 
             # Should select appropriate OpenAI models based on new preference order
-            assert extended_reasoning == "o3"  # O3 for extended reasoning
+            assert extended_reasoning == "gpt-5-codex"  # GPT-5-Codex prioritized for extended reasoning
             assert fast_response == "gpt-5"  # gpt-5 comes first in fast response preference
             assert balanced == "gpt-5"  # gpt-5 for balanced
 

tests/test_per_tool_model_defaults.py

@@ -95,8 +95,8 @@ class TestModelSelection:
             ModelProviderRegistry.register_provider(ProviderType.OPENAI, OpenAIModelProvider)
 
             model = ModelProviderRegistry.get_preferred_fallback_model(ToolModelCategory.EXTENDED_REASONING)
-            # OpenAI prefers o3 for extended reasoning
-            assert model == "o3"
+            # OpenAI prefers GPT-5-Codex for extended reasoning (coding tasks)
+            assert model == "gpt-5-codex"
 
     def test_extended_reasoning_with_gemini_only(self):
         """Test EXTENDED_REASONING prefers pro when only Gemini is available."""
@@ -192,7 +192,7 @@ class TestFlexibleModelSelection:
                 "env": {"OPENAI_API_KEY": "test-key"},
                 "provider_type": ProviderType.OPENAI,
                 "category": ToolModelCategory.EXTENDED_REASONING,
-                "expected": "o3",
+                "expected": "gpt-5-codex",  # GPT-5-Codex prioritized for coding tasks
             },
             # Case 2: Gemini provider for fast response
             {