Refactor runtime config architecture, expand model preset support, and harden LLM stream error observability

[StableLlama]

Description

• Move runtime user configuration from resources/config to data/config and remove legacy compatibility fallbacks.
• Add first-run bootstrap for machine/story/projects runtime config files so the WebUI starts without manual setup.
• Introduce tracked model preset database and schema, including Qwen 3.5 thinking/non-thinking presets and writing warnings.
• Expand machine model schema/UI/API support for advanced generation parameters (top_p, penalties, stop, seed, top_k, min_p, extra_body, preset metadata).
• Fix frontend model_name routing to prefer provider name for backend model resolution, ensuring model-scoped extra_body parameters are applied.
• Update sourcebook first-load behavior to return an empty list when no active project is selected.
• Improve LLM stream error handling to always persist upstream/internal failures to raw logs with non-empty fallback error details.
• Refresh backend/frontend contract coverage and docs for new preset endpoint and runtime config conventions.

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

How Has This Been Tested?

Please describe the tests that you ran to verify your changes.

• Backend tests (pytest)
• Frontend tests (npm run test)
• Manual verification

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules