feat: implement OAuth flow for vector search in CLI

[rach-kanc]

Summary

This PR finalizes the Vector Search & Embeddings OAuth implementation by exposing the new authentication capabilities to the developer CLI. Users can now securely simulate inbound vector queries using the --access-token flag, and configure dynamic outbound Bearer token acquisition for enterprise embedding models (like Azure OpenAI) using the --oauth-endpoint flags.

Protocol Alignment

• CAP — Context Access Protocol
• CCP — Context Contribution Protocol
• CRP — Context Rectification Protocol
• Identity Address
• Discovery
• Evidence Model
• Temporal Decay
• Provider Independence
• Documentation
• Community Health
• Other: ___

Changes Made

• src/cli.mjs: [MODIFY] Added new CLI arguments (--access-token, --oauth-client-id, --oauth-client-secret, --oauth-endpoint, --embedding-provider) and updated the query execution logic to dynamically instantiate the embedding service and perform a hybrid semantic search when OAuth options are provided.
• README.md: [MODIFY] Added a "Vector Search & Embeddings OAuth" section demonstrating how to use the new CLI flags for both inbound authorization and outbound authentication.

Testing

• Unit tests added/updated
• Manual testing done (Executed test suite locally to verify CLI flag parsing changes did not disrupt any of the 131 core tests)
• Integration tests pass

Checklist

• Code follows the repo's style guidelines
• Self-reviewed before opening PR
• No breaking changes to protocol-level interfaces (or documented in description)
• CHANGELOG updated if applicable

Linked Issue

Closes #161