fix: add anthropic to dev deps, format files

Author: echobt

pyproject.toml

@@ -6,7 +6,7 @@ requires-python = ">=3.11"
 dependencies = ["aiohttp>=3.9.0", "aiodocker>=0.22.0", "pydantic>=2.5.0", "pydantic-settings>=2.1.0", "litellm>=1.20.0", "typer>=0.9.0", "rich>=13.7.0", "aiosqlite>=0.19.0", "aiofiles>=23.2.0", "pyarrow>=15.0.0", "huggingface_hub>=0.20.0", "tenacity>=8.2.0", "structlog>=24.1.0"]
 
 [project.optional-dependencies]
-dev = ["pytest>=8.0.0", "pytest-asyncio>=0.23.0", "pytest-cov>=4.1.0", "mypy>=1.8.0", "ruff>=0.2.0"]
+dev = ["pytest>=8.0.0", "pytest-asyncio>=0.23.0", "pytest-cov>=4.1.0", "mypy>=1.8.0", "ruff>=0.2.0", "anthropic>=0.18.0"]
 
 [project.scripts]
 swe-forge = "swe_forge.__main__:main"

src/swe_forge/swe/models.py

@@ -101,8 +101,7 @@ class SweTask(BaseModel):
     fail_to_pass: list[str] = Field(default_factory=list)
     pass_to_pass: list[str] = Field(default_factory=list)
     install_config: dict[str, Any] = Field(
-        default_factory=dict,
-        description="Agentic-detected install configuration"
+        default_factory=dict, description="Agentic-detected install configuration"
     )
     meta: dict[str, str] = Field(default_factory=dict)
     prompt: str = ""
@@ -125,14 +124,14 @@ def validate_repo_name_field(cls, v: str) -> str:
 
     def is_install_ready(self) -> bool:
         """Check if install_config has been populated by agent.
-        
+
         Returns:
             True if agent has detected working install commands.
         """
         return bool(
-            self.install_config.get("install_commands") or
-            self.install_config.get("commands")
+            self.install_config.get("install_commands")
+            or self.install_config.get("commands")
         ) and self.install_config.get("validated", False)
-    
+
     def has_tests(self) -> bool:
         return bool(self.fail_to_pass or self.pass_to_pass)

tests/test_swe/test_agentic_config.py

@@ -12,7 +12,7 @@
 
 class TestRepositoryConfig:
     """Test RepositoryConfig dataclass."""
-    
+
     def test_config_creation(self):
         """Test basic config creation."""
         config = RepositoryConfig(
@@ -24,11 +24,11 @@ def test_config_creation(self):
             docker_image="python:3.11-slim",
             validated=True,
         )
-        
+
         assert config.python_version == "3.11"
         assert config.package_manager == "poetry"
         assert config.is_valid() is True
-    
+
     def test_config_invalid_missing_fields(self):
         """Test config validation with missing fields."""
         # Missing install_commands
@@ -40,7 +40,7 @@ def test_config_invalid_missing_fields(self):
             validated=True,
         )
         assert config1.is_valid() is False
-        
+
         # Missing test_command
         config2 = RepositoryConfig(
             python_version="3.11",
@@ -50,7 +50,7 @@ def test_config_invalid_missing_fields(self):
             validated=True,
         )
         assert config2.is_valid() is False
-    
+
     def test_config_not_validated(self):
         """Test config that wasn't validated."""
         config = RepositoryConfig(
@@ -60,14 +60,14 @@ def test_config_not_validated(self):
             test_command="pytest",
             validated=False,  # Not validated!
         )
-        
+
         # Should be invalid even if fields are filled
         assert config.is_valid() is False
 
 
 class TestSubmitConfigArgs:
     """Test SubmitConfigArgs validation."""
-    
+
     def test_submit_config_args(self):
         """Test SubmitConfigArgs creation."""
         args = SubmitConfigArgs(
@@ -77,23 +77,23 @@ def test_submit_config_args(self):
             test_command="pytest",
             test_framework="pytest",
         )
-        
+
         assert args.python_version == "3.11"
         assert args.package_manager == "poetry"
 
 
 class TestEmptyConfig:
     """Test that empty config is invalid."""
-    
+
     def test_empty_config_invalid(self):
         """Test that empty config is invalid."""
         config = RepositoryConfig()
-        
+
         # Should be invalid - no fields set
         assert config.is_valid() is False
         assert config.python_version == ""
         assert config.package_manager == ""
-        
+
     def test_config_with_validation_errors(self):
         """Test config with validation errors."""
         config = RepositoryConfig(
@@ -103,7 +103,7 @@ def test_config_with_validation_errors(self):
             test_command="",
             validated=False,
         )
-        
+
         assert config.is_valid() is False
         assert len(config.validation_errors) >= 0  # May have no errors explicitly
 

tests/test_swe/test_dataset_validator.py

@@ -15,7 +15,7 @@
 
 class TestValidationResult:
     """Test ValidationResult dataclass."""
-    
+
     def test_result_creation(self):
         """Test basic result creation."""
         result = ValidationResult(
@@ -25,11 +25,11 @@ def test_result_creation(self):
             install_commands_worked=["pip install -e ."],
             test_results={"base:pytest": False, "patch:pytest": True},
         )
-        
+
         assert result.passed is True
         assert result.phase == "complete"
         assert len(result.install_commands_worked) == 1
-    
+
     def test_result_to_dict(self):
         """Test result serialization."""
         result = ValidationResult(
@@ -39,34 +39,34 @@ def test_result_to_dict(self):
             error_message="Install failed",
             install_commands_failed=["pip install -e ."],
         )
-        
+
         data = result.to_dict()
-        
+
         assert data["task_id"] == "test-002"
         assert data["passed"] is False
         assert data["error_message"] == "Install failed"
-    
+
     def test_failed_install_tracking(self):
         """Test tracking of failed install commands."""
         result = ValidationResult(task_id="test-003", passed=False, phase="install")
-        
+
         result.install_commands_worked.append("apt-get update")
         result.install_commands_failed.append("pip install broken-package")
-        
+
         assert len(result.install_commands_worked) == 1
         assert len(result.install_commands_failed) == 1
 
 
 class TestDatasetValidationReport:
     """Test DatasetValidationReport."""
-    
+
     def test_empty_report(self):
         """Test empty report."""
         report = DatasetValidationReport()
-        
+
         assert report.total_tasks == 0
         assert report.pass_rate == 0.0
-    
+
     def test_report_with_results(self):
         """Test report with validation results."""
         report = DatasetValidationReport(
@@ -76,40 +76,45 @@ def test_report_with_results(self):
             validation_results=[
                 ValidationResult(task_id="t1", passed=True, phase="complete"),
                 ValidationResult(task_id="t2", passed=True, phase="complete"),
-                ValidationResult(task_id="t3", passed=False, phase="test_base", error_message="Test failed"),
+                ValidationResult(
+                    task_id="t3",
+                    passed=False,
+                    phase="test_base",
+                    error_message="Test failed",
+                ),
             ],
         )
-        
-        assert report.pass_rate == 2/3
+
+        assert report.pass_rate == 2 / 3
         assert report.failed_tasks == 1
-    
+
     def test_report_to_json(self):
         """Test report JSON serialization."""
         report = DatasetValidationReport(
             total_tasks=2,
             passed_tasks=1,
             failed_tasks=1,
         )
-        
+
         json_str = report.to_json()
         data = json.loads(json_str)
-        
+
         assert data["total_tasks"] == 2
         assert data["pass_rate"] == 0.5
 
 
 class TestPrintValidationReport:
     """Test print_validation_report function."""
-    
+
     def test_print_empty_report(self, capsys):
         """Test printing empty report."""
         report = DatasetValidationReport()
         print_validation_report(report)
-        
+
         captured = capsys.readouterr()
         assert "Total tasks: 0" in captured.out
         assert "Pass rate:" in captured.out
-    
+
     def test_print_report_with_results(self, capsys):
         """Test printing report with results."""
         report = DatasetValidationReport(
@@ -119,16 +124,16 @@ def test_print_report_with_results(self, capsys):
             validation_results=[
                 ValidationResult(task_id="t1", passed=True, phase="complete"),
                 ValidationResult(
-                    task_id="t2", 
-                    passed=False, 
+                    task_id="t2",
+                    passed=False,
                     phase="test_base",
-                    error_message="Test failed on base"
+                    error_message="Test failed on base",
                 ),
             ],
         )
-        
+
         print_validation_report(report)
-        
+
         captured = capsys.readouterr()
         assert "Passed: 3" in captured.out
         assert "Failed: 2" in captured.out
@@ -137,7 +142,7 @@ def test_print_report_with_results(self, capsys):
 
 class TestSweTaskInstallConfig:
     """Test SweTask with install_config field."""
-    
+
     def test_task_with_install_config(self):
         """Test task creation with install_config."""
         task = SweTask(
@@ -152,10 +157,10 @@ def test_task_with_install_config(self):
                 "validated": True,
             },
         )
-        
+
         assert task.install_config["python_version"] == "3.11"
         assert task.install_config["package_manager"] == "poetry"
-    
+
     def test_task_is_install_ready(self):
         """Test is_install_ready method."""
         # Not ready
@@ -164,7 +169,7 @@ def test_task_is_install_ready(self):
             install_config={},
         )
         assert task1.is_install_ready() is False
-        
+
         # Ready
         task2 = SweTask(
             id="test-002",