fix(studio): load complete live session history

[ehz0ah]

Description

Web Studio used the server's default 128,000-token context window while loading a session's complete history. A large current turn could therefore hide older completed live turns even though they remained in messages.jsonl.

This change requests the largest exactly representable JavaScript integer as the token budget only for complete-history hydration. Model-facing context calls keep the existing default.

Human Involvement

• A human participated in the implementation or review loop
• This PR was generated entirely by AI agents without human participation in the loop

Related Issue

External evidence: openai/codex#54051

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Refactoring (no functional changes)
• Performance improvement
• Test update

Changes Made

• Send an explicit complete-history token budget from fetchSessionMessages.
• Leave ordinary session-context requests unchanged.
• Verify the exact complete-history request in the existing API contract test.

Testing

• 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
• I have tested this on the following platforms:
  • Linux
  • macOS
  • Windows

Validation completed:

• 966 Web Studio tests passed across 133 files.
• Prettier and targeted ESLint checks passed.
• The production build passed.
• A source-pinned five-turn experiment improved visible history from one turn to all five, with no dropped or truncated messages.

Checklist

• My code follows the project's coding style
• I have performed a self-review of my 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
• Any dependent changes have been merged and published

Screenshots (if applicable)

Not applicable.

Additional Notes

The server's 128,000-token default and all model-context behavior remain unchanged.