What problem does it solve? Manually cycling through AI code review feedback—reading Greptile comments, fixing issues, pushing changes, and re-triggering reviews—consumes significant developer time across GitHub, GitLab, and Perforce workflows. ## Core Features & Use Cases - Automated Review-Fix Loop: Triggers Greptile reviews, polls check runs or pipelines for completion, fixes actionable comments, resolves threads, and re-pushes until a 5/5 confidence score with zero unresolved comments is reached. - Multi-Platform Support: Works with GitHub (gh CLI), GitLab (glab CLI), and Perforce (p4 CLI with shelved changelists), auto-detecting the platform from the environment. - Bounded Iteration: Caps cycles with a configurable --max-iterations flag (default 10) and reports remaining issues when the cap is hit. - Use Case: A developer opens a large PR that Greptile flags with 12 comments and a 3/5 score; greploop fixes each comment, resolves threads via GraphQL or the GitLab API, and repeats until the review is clean. ## Quick Start Run greploop on my current pull request and keep fixing Greptile's comments until it reaches a 5/5 confidence score.