feat: prompt for and configure GCP and BigQuery variables during installation

[drosales57]

This PR updates the installation workflows to prompt users for essential Google Cloud configuration variables during setup. These inputs are then dynamically injected into the plugin's .mcp.json configuration file, ensuring the various Data Cloud toolboxes (BigQuery, Spanner, AlloyDB, etc.) are pre-configured with the correct project, region, and location details upon installation.

Key Changes

• Interactive Installation Scripts (codex-install.sh & codex-install.ps1):

  • Added prompts to collect GCP Project ID, GCP Region, and BigQuery Location with clear, user-friendly descriptions.
  • Implemented logic to copy the local .mcp.json template to the installation directory and perform inline replacements of configuration placeholders with the user's actual values. (codex doesn't natively support env var expansion inside the json file so we replace values as part of the script)
  • Bash specific: Enhanced the read prompts with </dev/tty to ensure robust input handling even if the script is executed via a piped command or script runner.

• Configuration Templates (.claude-mcp.json & gemini-extension.json):

  • Updated environment variable definitions across all toolboxes (including BigQuery, Spanner, AlloyDB, Cloud SQL, Dataplex, Dataproc, and Serverless Spark) to use placeholders instead of empty strings.
  • for Claude code, we do this using the user configuration support in plugins
  • for gemini CLI, we use extension settings