docs: clarify Field description handling for variadic tool parameters

[subhashpolisetti]

Summary

This pull request clarifies how a Pydantic Field description is handled for variadic tool parameters.

The section is titled "Constraining and describing arguments with Pydantic Field" and states that Field adds constraints and descriptions to tool arguments. The variadic paragraph then mentions only constraints, so the wording reads as though Field(description=...) also reaches a variadic parameter. It does not: on v0.22.2, *scores: Annotated[int, Field(description="Exam scores", ge=0, le=100)] advertises the constraints on each collected value and no parameter description.

The added sentence states that the SDK ignores a description passed to Field(...) on a variadic parameter and names the two supported ways that description is taken, following the guidance given when #4951 was closed.

Test plan

• Verified on v0.22.2 that *scores: Annotated[int, "Exam scores", Field(ge=0, le=100)] and a docstring entry alongside Field(ge=0, le=100) each produce the parameter description together with the per-value constraints, and that Field(description=...) alone produces no description.
• make build-docs completes with no errors and no new warnings. Generated pages under docs/ja, docs/ko, and docs/zh are untouched.
• This is a docs-only change, so the SDK verification stack does not apply; make build-docs is the applicable check.

Checks

• I've added new tests, if relevant
• I've run .agents/skills/code-change-verification/scripts/run.sh
• I've confirmed all verification steps pass
• If using Codex, I've run /review before submitting this PR

[github-actions]

This PR is stale because it has been open for 10 days with no activity.

[jbeckwith-oai]

Thank you, @subhashpolisetti, for identifying this documentation gap, writing the original clarification and example, and verifying the supported alternatives.

We have carried your contribution forward in #5223, with the description guidance in its own paragraph and explicit docstring precedence. The replacement PR credits you prominently, and its commit includes you as a co-author using the authorship from your original commit.

Closing this draft in favor of #5223 so we can take the clarification through review there. Thank you again for the investigation and contribution.