Finishing touches

[svetstoykov]

This pull request introduces initial release documentation, CI/CD improvements, and better sample configuration management for the TokenGuard project. It adds a release validation workflow, a changelog, and a detailed release process guide. Additionally, it improves developer experience by providing example configuration files for samples, updating documentation to clarify local secret handling, and refining explanations of token counting and compaction behavior.

Release process and CI/CD automation:

• Added .github/workflows/release-validation.yml to enforce pre-release validation, including restore, build, test, pack, and file-integrity checks for all public packages.
• Added docs/release-process.md as the authoritative runbook for NuGet package publishing, detailing prerequisites, validation gates, versioning, and publication order.
• Added CHANGELOG.md to document all notable changes and establish versioning and release notes standards.

Sample configuration and documentation improvements:

• Added example config files (appsettings.Development.example.json) for both samples/Codexplorer and samples/Codexplorer.Automation, and updated README.md files to instruct users to copy these for local secrets, keeping credentials out of source control. [1] [2] [3] [4] [5] [6]
• Updated documentation to clarify local build and run instructions, automation setup, and the importance of not committing secrets. [1] [2] [3] [4] [5]

Core documentation and feature clarifications:

• Improved explanations in README.md and deep-dive docs regarding token counting, provider anchoring, compaction thresholds, and configuration surface—highlighting that EstimatedTokenCounter is always used by default and provider token anchoring is optional. [1] [2] [3] [4] [5] [6] [7] [8]

Minor project structure cleanup:

• Removed the obsolete consolidated-audit-report.md from the TokenGuard.sln solution file.

Sample configuration defaults:

• Updated Codexplorer.Automation and Codexplorer sample appsettings.json files to use empty or placeholder paths and keys, ensuring no sensitive or user-specific data is committed. [1] [2]

These changes collectively lay the foundation for public release, enforce robust validation, document the release process, and improve onboarding for new users and contributors.