Review and update INPUT_CONTENT_LIMIT of summarize_review prompt
Everyone can contribute. Help move this issue forward while earning points, leveling up and collecting rewards.
Background
@dskim_gitlab brought up in !167364 (comment 2157247043) that the input content limiting logic we have in summarize review feature is OpenAI specific. We should see if this still applies with the current model we are using and update it if it no longer applies.
Based on https://docs.anthropic.com/en/docs/about-claude/models#model-comparison-table, Claude 3.5 Sonnet has 200k context window.
Proposal
Review the INPUT_CONTENT_LIMIT in https://gitlab.com/gitlab-org/gitlab/-/blob/636fdddf2c90703702da94dcb9ba2a463487ca81/ee/lib/gitlab/llm/templates/summarize_review.rb and https://gitlab.com/gitlab-org/gitlab/-/blob/b7be65c580ff9252914be56dd8fe02640414e127/ee/lib/gitlab/llm/ai_gateway/completions/summarize_review.rb and see if that still applies to current model we are using (we are currently using Claude 3.5 Sonnet).
If we can utilize the bigger context window of the new model, update the INPUT_CONTENT_LIMIT.