adk-samples agent: Agent Optimizer and Evaluation

[enriquekalven]

🚀 PR Description: Agent Optimizer & Evaluation Agent

📋 Overview
This PR introduces the Agent Optimizer, a developer tool built with the Google Agent Development Kit (ADK). It is designed to help developers analyze, optimize, and evaluate their agent implementations based on context engineering principles and best practices. 🛠️

✨ Key Features

• 🧠 Context Engineering Analysis: Automatically evaluates agents against the "5 Pillars" framework (Offload, Reduce, Retrieve, Isolate, Cache).
• 🌿 Evergreen Knowledge: Dynamically searches for and ingests the latest ADK and Agent Engine features to ensure recommendations are up-to-date.
• 🛡️ Gemini Code Assist Integration: Leverages Gemini Code Assist for deep architectural and security reviews.
• 🔌 Model Context Protocol (MCP): Supports connecting to external developer tools via MCP for enhanced implementation context.
• 📊 Adaptive Rubric Generation: Automatically generates customized evaluation rubrics tailored to an agent's specific goals and requirements.

🛠️ Changes
📂 [NEW] python/agents/agent-optimizer/: Complete implementation of the Agent Optimizer.

• agent_optimizer/: Core logic, prompts, and analysis tools.
• samples/monolith/: A reference implementation of a "Customer Service" monolith agent to demonstrate optimization patterns.
• Makefile: Automation for deployment to Vertex AI Agent Engine and registration with Gemini Enterprise.
• README.md: Comprehensive documentation and setup guide.
• pyproject.toml: Modern Python project configuration (pdm/uv compatible).
• 🌐 Environment Support: Added .env.example and .gitignore for standard development workflows.

✅ Verification

• 🧪 Successfully ran adk run locally to analyze sample agents.
• 🚢 Verified make deploy workflow is compatible with Agent Engine.
• 💻 Verified adk web playground functionality.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.