fix: reduce pedantic reviews, surface human replies, raise diff/token limits

[chsrimanaswi]

fix: prompt change for review, surface human replies, raise diff/token limits

Description

• Scoped best practices feedback to actual runtime context.
• Re-raise logic now fetches human replies to bot inline comments and passes them as humanResponse so Claude can weigh developer explanations before deciding to re-raise
• Diff character limit raised from 100k to 500k (extracted to DIFF_CHAR_LIMIT constant)
• max_tokens raised from 8192 to 32768 to prevent truncated JSON responses on large diffs

Related Issue

Motivation and Context

The reviewer was raising overly pedantic issues (e.g. singleton test isolation concerns on one-shot scripts, version compatibility for unsupported Node versions). Additionally, when a developer replied to a bot comment explaining their reasoning, the bot would ignore that response and blindly re-raise the same issue on the next review trigger.

How Has This Been Tested?

Manually reviewed script logic. End-to-end validation via a live PR trigger pending.

Screenshots (if appropriate):

Types of changes

• 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 change)

Checklist:

• I have signed the Adobe Open Source CLA.
• My code follows the code style of this project.
• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• I have read the CONTRIBUTING document.
• I have added tests to cover my changes.
• All new and existing tests passed.

[chsrimanaswi]

tested here:
adobe/aio-lib-state#253 (review)