[Repo Assist] Add TypeInference documentation page for type inference and missing values

[github-actions]

🤖 This is an automated PR from Repo Assist, created in response to the request on issue #347.

Closes #347

What this PR does

Adds a new documentation page docs/library/TypeInference.fsx that comprehensively explains type inference rules and missing-value handling across all FSharp.Data type providers. The page is linked from docs/index.md.

Content

The new page covers:

• Numeric type hierarchy — how int, int64, decimal, and float are chosen and how types are promoted when samples mix values
• Boolean inference in CSV — which string values produce bool
• Date and time inference — DateTime, DateTimeOffset, DateOnly, TimeOnly, and the role of PreferDateOnly
• Missing values in JSON — how absent properties and null values both become option(T), with a note on design-time vs runtime risks
• Missing values in CSV — the default missing-value strings (NaN, NA, etc.), and a comparison table showing the default type vs. PreferOptionals=true for each base type
• Missing values in XML — how absent attributes and elements become option(T)
• Heterogeneous types — how JSON values that can be multiple types are exposed via multiple optional accessors
• Design-time vs runtime behaviour — key caveats about what happens when runtime data differs from the sample
• Summary table of all inference-control parameters (PreferOptionals, AssumeMissingValues, MissingValues, InferRows, SampleIsList, PreferDateOnly, InferenceMode, Schema) with the providers each applies to

Test Status

Build: ✅ passed (dotnet run --project build/build.fsproj -t Build)
Fantomas format check: ✅ passed (dotnet fantomas --check docs/library/TypeInference.fsx)
No executable code was changed; this is a documentation-only addition.

Generated by Repo Assist for issue #347

To install this agentic workflow, run

gh aw add githubnext/agentics/workflows/repo-assist.md@afb00b92a9514fee9a14c583f059a03d05738f70