diff --git a/checks.sh b/checks.sh
index 39fb26e7..36124e25 100755
--- a/checks.sh
+++ b/checks.sh
@@ -15,8 +15,8 @@ headerEnd=" ===\033[0m\n"
 
 echo "${headerStart}Checking Python: Ruff, format, check${headerEnd}"
 # I is import sorting
-ruff check --select I
-ruff format --check
+uvx  ruff check --select I
+uvx ruff format --check .
 
 echo "${headerStart}Checking for Misspellings${headerEnd}"
 find . -type f | grep -v "/node_modules/" | grep  -v "/\." | grep -v "/dist/" | grep -v "/desktop/build/" | xargs misspell -error
diff --git a/libs/core/kiln_ai/adapters/prompt_builders.py b/libs/core/kiln_ai/adapters/prompt_builders.py
index e595bd71..04d5af0c 100644
--- a/libs/core/kiln_ai/adapters/prompt_builders.py
+++ b/libs/core/kiln_ai/adapters/prompt_builders.py
@@ -103,7 +103,7 @@ def build_prompt(self) -> str:
             )
             # iterate requirements, formatting them in numbereed list like 1) task.instruction\n2)...
             for i, requirement in enumerate(self.task.requirements):
-                base_prompt += f"{i+1}) {requirement.instruction}\n"
+                base_prompt += f"{i + 1}) {requirement.instruction}\n"
 
         return base_prompt
 
@@ -126,12 +126,12 @@ def build_prompt(self) -> str:
         Returns:
             str: The constructed prompt string with examples.
         """
-        base_prompt = f"# Instruction\n\n{ self.task.instruction }\n\n"
+        base_prompt = f"# Instruction\n\n{self.task.instruction}\n\n"
 
         if len(self.task.requirements) > 0:
             base_prompt += "# Requirements\n\nYour response should respect the following requirements:\n"
             for i, requirement in enumerate(self.task.requirements):
-                base_prompt += f"{i+1}) {requirement.instruction}\n"
+                base_prompt += f"{i + 1}) {requirement.instruction}\n"
             base_prompt += "\n"
 
         valid_examples = self.collect_examples()
@@ -148,7 +148,7 @@ def build_prompt(self) -> str:
     def prompt_section_for_example(self, index: int, example: TaskRun) -> str:
         # Prefer repaired output if it exists, otherwise use the regular output
         output = example.repaired_output or example.output
-        return f"## Example {index+1}\n\nInput: {example.input}\nOutput: {output.output}\n\n"
+        return f"## Example {index + 1}\n\nInput: {example.input}\nOutput: {output.output}\n\n"
 
     def collect_examples(self) -> list[TaskRun]:
         valid_examples: list[TaskRun] = []
@@ -206,7 +206,7 @@ def prompt_section_for_example(self, index: int, example: TaskRun) -> str:
         ):
             return super().prompt_section_for_example(index, example)
 
-        prompt_section = f"## Example {index+1}\n\nInput: {example.input}\n\n"
+        prompt_section = f"## Example {index + 1}\n\nInput: {example.input}\n\n"
         prompt_section += (
             f"Initial Output Which Was Insufficient: {example.output.output}\n\n"
         )
diff --git a/libs/core/kiln_ai/adapters/test_prompt_builders.py b/libs/core/kiln_ai/adapters/test_prompt_builders.py
index f6d55736..f60c6993 100644
--- a/libs/core/kiln_ai/adapters/test_prompt_builders.py
+++ b/libs/core/kiln_ai/adapters/test_prompt_builders.py
@@ -241,14 +241,14 @@ def test_few_shot_prompt_builder(tmp_path):
     # Create 6 examples (2 repaired, 4 high-quality)
     for i in range(6):
         run = TaskRun(
-            input=f'{{"subject": "Subject {i+1}"}}',
+            input=f'{{"subject": "Subject {i + 1}"}}',
             input_source=DataSource(
                 type=DataSourceType.human,
                 properties={"created_by": "john_doe"},
             ),
             parent=task,
             output=TaskOutput(
-                output=f'{{"joke": "Joke Initial Output {i+1}"}}',
+                output=f'{{"joke": "Joke Initial Output {i + 1}"}}',
                 source=DataSource(
                     type=DataSourceType.human,
                     properties={"created_by": "john_doe"},
@@ -262,7 +262,7 @@ def test_few_shot_prompt_builder(tmp_path):
                 update={
                     "repair_instructions": "Fix the joke",
                     "repaired_output": TaskOutput(
-                        output=f'{{"joke": "Repaired Joke {i+1}"}}',
+                        output=f'{{"joke": "Repaired Joke {i + 1}"}}',
                         source=DataSource(
                             type=DataSourceType.human,
                             properties={"created_by": "jane_doe"},
@@ -335,6 +335,9 @@ def test_prompt_builder_from_ui_name(task_with_examples):
     with pytest.raises(ValueError, match="Unknown prompt builder: invalid_name"):
         prompt_builder_from_ui_name("invalid_name", task)
 
+    with pytest.raises(ValueError, match="Prompt ID not found: 123"):
+        prompt_builder_from_ui_name("id::123", task)
+
 
 def test_example_count():
     assert FewShotPromptBuilder.example_count() == 4
diff --git a/libs/server/kiln_server/test_run_api.py b/libs/server/kiln_server/test_run_api.py
index 9df600e2..ace8e13e 100644
--- a/libs/server/kiln_server/test_run_api.py
+++ b/libs/server/kiln_server/test_run_api.py
@@ -463,12 +463,12 @@ async def test_update_run(client, tmp_path):
                 json=case["updates"],
             )
 
-            assert (
-                response.status_code == case["expected_status"]
-            ), f"Failed on case: {case['name']}"
-            assert (
-                response.json()["message"] == case["expected_detail"]
-            ), f"Failed on case: {case['name']}"
+            assert response.status_code == case["expected_status"], (
+                f"Failed on case: {case['name']}"
+            )
+            assert response.json()["message"] == case["expected_detail"], (
+                f"Failed on case: {case['name']}"
+            )
 
 
 @pytest.fixture