Add logo, improve README and SDK README

- Add VerifyWise logo with light/dark mode support
- Center-aligned header with tagline and badges
- Rewrite SDK README with practical examples for every major feature
- Add dark-mode SVG variant for GitHub rendering

Made-with: Cursor

Author: EfeAcar6431

.github/assets/verifywise-logo-dark.svg

@@ -1,12 +1,29 @@
-# VerifyWise LLM Evaluation Action
+<p align="center">
+  <a href="https://verifywise.ai">
+    <picture>
+      <source media="(prefers-color-scheme: dark)" srcset="https://raw.githubusercontent.com/verifywise-ai/verifywise-eval-action/main/.github/assets/verifywise-logo-dark.svg">
+      <source media="(prefers-color-scheme: light)" srcset="https://raw.githubusercontent.com/verifywise-ai/verifywise-eval-action/main/.github/assets/verifywise-logo.svg">
+      <img alt="VerifyWise" src="https://raw.githubusercontent.com/verifywise-ai/verifywise-eval-action/main/.github/assets/verifywise-logo.svg" width="300">
+    </picture>
+  </a>
+</p>
+
+<h1 align="center">LLM Evaluation Action</h1>
+
+<p align="center">
+  <strong>Automatically evaluate your LLMs on every pull request.</strong><br>
+  Gate merges on correctness, faithfulness, hallucination, toxicity, and bias — enforced in CI.
+</p>
+
+<p align="center">
+  <a href="https://github.com/verifywise-ai/verifywise-eval-action"><img src="https://img.shields.io/badge/GitHub%20Action-VerifyWise%20Eval-green?logo=github" alt="GitHub Action"></a>
+  <a href="LICENSE"><img src="https://img.shields.io/badge/License-Apache%202.0-blue.svg" alt="License"></a>
+  <a href="https://pypi.org/project/verifywise/"><img src="https://img.shields.io/pypi/v/verifywise?label=SDK" alt="Python SDK"></a>
+</p>
 
-[![GitHub Action](https://img.shields.io/badge/GitHub%20Action-VerifyWise%20Eval-green?logo=github)](https://github.com/verifywise-ai/verifywise-eval-action)
-[![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE)
-[![Python SDK](https://img.shields.io/pypi/v/verifywise?label=SDK)](https://pypi.org/project/verifywise/)
-
-**Automatically evaluate your LLMs on every pull request.** This action connects to [VerifyWise](https://verifywise.ai) — the open-source AI governance platform — to run your models against curated test datasets and block merges when quality drops below your standards.
+---
 
-> Think of it as unit tests for your LLM: correctness, faithfulness, hallucination, toxicity, bias — measured on every code change, enforced in CI.
+This action connects to [VerifyWise](https://verifywise.ai) — the open-source AI governance platform — to run your models against curated test datasets and block merges when quality drops below your standards.
 
 ---