chore: Reformatted the image-scoring sample by ensuring black, isort, and flake8 pass

python/agents/image-scoring/eval/test_eval.py

@@ -1,7 +1,7 @@
-
 """Basic evalualtion for Image Scoring."""
 
 import pathlib
+
 import dotenv
 import pytest
 from google.adk.evaluation import AgentEvaluator

python/agents/image-scoring/image_scoring/__init__.py

@@ -1 +1 @@
-from . import agent
+from . import agent  # noqa: F401

python/agents/image-scoring/image_scoring/agent.py

@@ -1,14 +1,17 @@
-import datetime, uuid
-from zoneinfo import ZoneInfo
+import datetime
 import os
+import uuid
+from zoneinfo import ZoneInfo
+
 import google.auth
-from .sub_agents.prompt import image_generation_prompt_agent 
-from .sub_agents.image import image_generation_agent 
-from .sub_agents.scoring import scoring_images_prompt 
-from .checker_agent import checker_agent_instance
-from google.adk.agents import SequentialAgent, LoopAgent
+from google.adk.agents import LoopAgent, SequentialAgent
 from google.adk.agents.callback_context import CallbackContext
 
+from .checker_agent import checker_agent_instance
+from .sub_agents.image import image_generation_agent
+from .sub_agents.prompt import image_generation_prompt_agent
+from .sub_agents.scoring import scoring_images_prompt
+
 # To use AI Studio credentials:
 # 1. Create a .env file in the /app directory with:
 #    GOOGLE_GENAI_USE_VERTEXAI=FALSE
@@ -43,7 +46,6 @@ def set_session(callback_context: CallbackContext):
 
 image_generation_scoring_agent = SequentialAgent(
     name="image_generation_scoring_agent",
-
     description=(
         """
         Analyzes a input text and creates the image generation prompt, generates the relevant images with imagen3 and scores the images."
@@ -52,7 +54,11 @@ def set_session(callback_context: CallbackContext):
         3. Invoke the scoring_images_prompt agent to score the images
             """
     ),
-    sub_agents=[image_generation_prompt_agent, image_generation_agent, scoring_images_prompt],
+    sub_agents=[
+        image_generation_prompt_agent,
+        image_generation_agent,
+        scoring_images_prompt,
+    ],
 )
 
 

python/agents/image-scoring/image_scoring/checker_agent.py

@@ -1,9 +1,9 @@
 from google.adk.agents import Agent
+
 from . import config
 from .prompt import CHECKER_PROMPT
 from .tools.loop_condition_tool import check_tool_condition
 
-
 # This agent is responsible for checking conditions and validating the scoring process
 # It uses the check_tool_condition tool to evaluate whether the scoring process should continue
 # The agent's output is stored in the "checker_output" key

python/agents/image-scoring/image_scoring/prompt.py

@@ -1,5 +1,4 @@
-CHECKER_PROMPT = """
-You are an agent to evaluate the quality of image based on the total_score of the image 
+CHECKER_PROMPT = """You are an agent to evaluate the quality of image based on the total_score of the image
 generation.
 
 1. Invoke the `image_generation_scoring_agent` first to generate images and score the images.

python/agents/image-scoring/image_scoring/sub_agents/image/__init__.py

@@ -1 +1 @@
-from .imagen_agent import image_generation_agent
+from .imagen_agent import image_generation_agent  # noqa: F401

python/agents/image-scoring/image_scoring/sub_agents/image/imagen_agent.py

@@ -1,7 +1,8 @@
-from .prompt import IMAGEGEN_PROMPT
 from google.adk.agents import Agent
-from .tools.image_generation_tool import generate_images
+
 from ... import config
+from .prompt import IMAGEGEN_PROMPT
+from .tools.image_generation_tool import generate_images
 
 image_generation_agent = Agent(
     name="image_generation_agent",

python/agents/image-scoring/image_scoring/sub_agents/image/prompt.py

@@ -1,4 +1,4 @@
 IMAGEGEN_PROMPT = """
-Your job is to invoke the 'generate_images' tool by passing the `image generation prompt` provided 
+Your job is to invoke the 'generate_images' tool by passing the `image generation prompt` provided
 to you as a parameter .
 """

python/agents/image-scoring/image_scoring/sub_agents/image/tools/image_generation_tool.py

@@ -1,14 +1,15 @@
-from datetime import datetime
 import os
+from datetime import datetime
+
 from google import genai
-from google.genai import types
 from google.adk.tools import ToolContext
 from google.cloud import storage
-from .... import config
+from google.genai import types
 
+from .... import config
 
 client = genai.Client(
-    vertexai=True, project=os.environ.get("GOOGLE_CLOUD_PROJECT"),location="global"
+    vertexai=True, project=os.environ.get("GOOGLE_CLOUD_PROJECT"), location="global"
 )
 
 
@@ -26,13 +27,12 @@ async def generate_images(imagen_prompt: str, tool_context: ToolContext):
                 person_generation="allow_adult",
             ),
         )
-        generated_image_paths = []
         if response.generated_images is not None:
             for generated_image in response.generated_images:
                 # Get the image bytes
                 image_bytes = generated_image.image.image_bytes
                 counter = str(tool_context.state.get("loop_iteration", 0))
-                artifact_name = f"generated_image_" + counter + ".png"
+                artifact_name = f"generated_image_{counter}.png"
                 # call save to gcs function
                 if config.GCS_BUCKET_NAME:
                     save_to_gcs(tool_context, image_bytes, artifact_name, counter)
@@ -60,8 +60,7 @@ async def generate_images(imagen_prompt: str, tool_context: ToolContext):
             }
 
     except Exception as e:
-
-        return {"status": "error", "message": "No images generated.  {e}"}
+        return {"status": "error", "message": f"No images generated. {e}"}
 
 
 def save_to_gcs(tool_context: ToolContext, image_bytes, filename: str, counter: str):

python/agents/image-scoring/image_scoring/sub_agents/prompt/__init__.py

@@ -1 +1 @@
-from .prompt_agent import image_generation_prompt_agent
+from .prompt_agent import image_generation_prompt_agent  # noqa: F401

python/agents/image-scoring/image_scoring/sub_agents/prompt/prompt.py

@@ -1,23 +1,24 @@
 PROMPT = """
- Your primary objective: Transform the input text into a pair of highly optimized prompts—one positive and 
+ Your primary objective: Transform the input text into a pair of highly optimized prompts—one positive and
  one negative—specifically designed for generating a visually compelling,
  rule-compliant lockscreen image using the Imagen3 text-to-image model (provided by Google/GCP).
-    Critical First Step: Before constructing any prompts, you must first analyze the 
+    Critical First Step: Before constructing any prompts, you must first analyze the
     input text to identify or conceptualize a primary subject. This subject MUST:
-    1. Be very much related to the input text presented. The viewer should  
-     feel that the generated image of that subject is conveying 
+    1. Be very much related to the input text presented. The viewer should
+     feel that the generated image of that subject is conveying
     what he/she is reading from that new article.
-    2. It should not  violate any content restrictions (especially regarding humans, 
+    2. It should not  violate any content restrictions (especially regarding humans,
     politics, religion, etc.).
     3. Describe in detail on what we would like to represent around the primary subject,
-      as-in, paint a complete picture. 
-    This chosen subject will be the cornerstone of your "Image Generation Prompt". 
-
-    Invoke the 'get_policy_text' tool to obtain the 'policy_text'. The 'policy_text' 
+      as-in, paint a complete picture.
+
+    This chosen subject will be the cornerstone of your "Image Generation Prompt".
+
+    Invoke the 'get_policy_text' tool to obtain the 'policy_text'. The 'policy_text'
     defines the rules for the image generation.
     The image also should comply with rules defined in the 'policy_text'.
-    
-    Negative Prompt: Generate a negative prompt to ensure the image does not 
+
+    Negative Prompt: Generate a negative prompt to ensure the image does not
     violate the rules defined in the 'policy_text'.
 
     """

python/agents/image-scoring/image_scoring/sub_agents/prompt/prompt_agent.py

@@ -1,7 +1,8 @@
-from ... import config
 from google.adk.agents import Agent
-from .prompt import PROMPT
+
+from ... import config
 from ..tools.fetch_policy_tool import get_policy
+from .prompt import PROMPT
 
 image_generation_prompt_agent = Agent(
     name="image_generation_prompt_agent",

python/agents/image-scoring/image_scoring/sub_agents/scoring/__init__.py

@@ -1 +1 @@
-from .scoring_agent import scoring_images_prompt
+from .scoring_agent import scoring_images_prompt  # noqa: F401

python/agents/image-scoring/image_scoring/sub_agents/scoring/prompt.py

@@ -55,15 +55,13 @@
           }\
         }"
 
-
         "Do not validate the JSON structure itself; only use its content for scoring rules. "
         "5. Compute the total_score by adding each individual score point for each rule in the JSON "
         "6. Invoke the set_score tool and pass the total_score. "
 
-
         "OUTPUT JSON FORMAT SPECIFICATION:\n"
         "The JSON object MUST have exactly two top-level keys:"
         "  - 'total_score': Iterate through each individual score element in the json and add those to arrive at total_score. "
         "  - 'scores': The existing rules json with a score attribute assigned to each rule and a reason attribute"
-  
+
 """

python/agents/image-scoring/image_scoring/sub_agents/scoring/scoring_agent.py

@@ -1,10 +1,10 @@
 from google.adk.agents import Agent
+
 from ... import config
 from ..tools.fetch_policy_tool import get_policy
+from .prompt import SCORING_PROMPT
 from .tools.get_images_tool import get_image
 from .tools.set_score_tool import set_score
-from .prompt import SCORING_PROMPT
-
 
 scoring_images_prompt = Agent(
     name="scoring_images_prompt",

python/agents/image-scoring/image_scoring/sub_agents/scoring/tools/get_images_tool.py

@@ -3,20 +3,18 @@
 
 async def get_image(tool_context: ToolContext):
     try:
-        
+
         artifact_name = (
-            f"generated_image_" + str(tool_context.state.get("loop_iteration", 0)) + ".png"
+            f"generated_image_{tool_context.state.get('loop_iteration', 0)}.png"
         )
-        artifact = await tool_context.load_artifact(artifact_name)
-
-
+        await tool_context.load_artifact(artifact_name)
 
         return {
             "status": "success",
-            "message": f"Image artifact {artifact_name} successfully loaded."
+            "message": f"Image artifact {artifact_name} successfully loaded.",
         }
     except Exception as e:
         return {
             "status": "error",
-            "message": f"Error loading artifact {artifact_name}: {str(e)}"
+            "message": f"Error loading artifact {artifact_name}: {str(e)}",
         }

python/agents/image-scoring/image_scoring/tools/loop_condition_tool.py

@@ -1,11 +1,10 @@
-from google.adk.tools import ToolContext, FunctionTool
-from .. import config
+from google.adk.tools import FunctionTool, ToolContext
 
+from .. import config
 
 
 def check_condition_and_escalate_tool(tool_context: ToolContext) -> dict:
     """Checks the loop termination condition and escalates if met or max count reached."""
-
 
     # Increment loop iteration count using state
     current_loop_count = tool_context.state.get("loop_iteration", 0)
@@ -34,14 +33,10 @@ def check_condition_and_escalate_tool(tool_context: ToolContext) -> dict:
         tool_context.actions.escalate = True
         response_message += "Max iterations reached, stopping loop."
     else:
-        print(
-            "  Condition not met and max iterations not reached. Loop will continue."
-        )
+        print("  Condition not met and max iterations not reached. Loop will continue.")
         response_message += "Loop continues."
 
     return {"status": "Evaluated scoring condition", "message": response_message}
 
 
-
-
 check_tool_condition = FunctionTool(func=check_condition_and_escalate_tool)