▸case-01 I just finished building a feature on my branch with an AI assistant, but it added a lot of unnecessary inline Python imports and redundant comments. Can you clean up my changes compared to the main branch and provide a brief summary of what you fixed? | fail→fail | 5,217 | 4,457 | -15% | 1 | 1 | 0% | 827 | 326 | -61% | 0 | 0 | — |
▸case-02 Hey, please review the unmerged commits on this branch and strip out any AI-generated noise like excessive try-catch blocks or type casts to 'any'. Make sure the updated code matches our project style and give me a short overview of what was updated. | fail→fail | 4,473 | 5,103 | +14% | 1 | 1 | 0% | 552 | 364 | -34% | 0 | 0 | — |
▸case-03 Could you inspect the diff between my current branch and main to prune low-quality AI code artifacts and extraneous defensive checks? Give me a concise 2-3 sentence report of the adjustments made once you're done. | fail→fail | 4,766 | 4,843 | +2% | 1 | 1 | 0% | 212 | 331 | +56% | 0 | 0 | — |
▸case-04 In our Python web service repo `user_service`, an LLM generated code that places `import json` and `import requests` inside function bodies to avoid global namespace clutter. Inspect our feature branch relative to `main` and clean it up. | fail→fail | 4,874 | 5,732 | +18% | 1 | 1 | 0% | 195 | 562 | +188% | 0 | 0 | — |
▸case-05 Our TypeScript backend `auth-service` has several newly added lines where values are cast as `as any` or `(<any>val)` because the LLM had trouble matching interfaces. Clean up these type escape hatches on the current branch compared to `main` and report the result. | fail→fail | 3,402 | 5,664 | +66% | 1 | 1 | 0% | 333 | 353 | +6% | 0 | 0 | — |
▸case-06 In our payment processing module `checkout.ts`, the order processing pipeline already validates `order.id` upstream in middleware. An AI coder added redundant `if (!order.id) throw new Error(...)` and `try/catch` wrappers around every line. Clean up these changes on the feature branch compared to `main`. | fail→fail | 7,473 | 4,914 | -34% | 1 | 1 | 0% | 1,324 | 350 | -74% | 0 | 0 | — |
▸case-07 I need to review changes on my feature branch `feature/billing` before opening a pull request to `main`. What exact git command should be used to inspect the changes introduced on this branch compared to `main`? | fail→pass | 6,281 | 9,942 | +58% | 1 | 1 | 0% | 969 | 1,310 | +35% | 0 | 0 | — |
▸case-08 The AI assistant left comments like `// Increment i by 1` and `// Call function to fetch user` across `user_controller.go` on this feature branch. Clean up the branch changes relative to `main` and summarize what was changed. | fail→fail | 4,367 | 3,804 | -13% | 1 | 1 | 0% | 516 | 345 | -33% | 0 | 0 | — |
▸case-09 In `data_pipeline.py`, after completing the code deslop review process against `main`, what is the target length for the final summary report? | fail→pass | 20,269 | 1,433 | -93% | 1 | 1 | 0% | 1,266 | 347 | -73% | 0 | 0 | — |
▸case-10 An AI wrote a data transformation script `etl_processor.py` on my branch and put `import pandas as pd` inside `process_row()`. I want to clean this up against `main` while preserving the actual ETL logic. Clean up the file. | fail→fail | 8,492 | 3,783 | -55% | 1 | 1 | 0% | 1,291 | 351 | -73% | 0 | 0 | — |
▸case-11 In `internal_analytics.rs`, function `aggregate_metrics` is only called by an internal worker that guarantees non-null initialized structs. An AI added nested `match` and `unwrap_or_default` fallback blocks for internal fields. Clean up this AI noise against `main`. | fail→fail | 7,209 | 4,851 | -33% | 1 | 1 | 0% | 1,285 | 345 | -73% | 0 | 0 | — |
▸case-12 An LLM introduced mixed single quotes and double quotes, as well as 2-space indentation alongside 4-space indentation in `config_parser.js` on my feature branch. Clean up the diff against `main`. | fail→fail | 10,949 | 4,963 | -55% | 1 | 1 | 0% | 1,812 | 380 | -79% | 0 | 0 | — |
▸case-13 In `api_client.ts`, an LLM cast response payload objects with `(res as any).data.items` to bypass strict TypeScript interface errors on the API response. Clean up these slop patterns on the feature branch compared to `main`. | fail→fail | 15,203 | 6,388 | -58% | 1 | 1 | 0% | 1,679 | 331 | -80% | 0 | 0 | — |
▸case-14 My feature branch contains changes across `server.py`, `utils.ts`, and `handler.go`. Each file has AI-generated noise like inline imports, `any` casts, and verbose self-explanatory comments. Run the cleanup process across the branch compared to `main`. | fail→fail | 7,404 | 4,731 | -36% | 1 | 1 | 0% | 1,144 | 318 | -72% | 0 | 0 | — |
▸case-15 When checking changes on a feature branch against `main` to find AI slop introduced strictly on this branch (excluding new commits added to `main` since branching), should I run `git diff main..HEAD` or `git diff main...HEAD`? | pass→pass | 6,357 | 3,143 | -51% | 1 | 1 | 0% | 1,114 | 652 | -41% | 0 | 0 | — |
▸case-16 In `cli_runner.py`, `import sys` and `import os` are called inside `def run_command()`. Clean up the code compared to `main` to adhere to clean Python standards. | fail→fail | 5,195 | 5,321 | +2% | 1 | 1 | 0% | 823 | 341 | -59% | 0 | 0 | — |
▸case-17 I have a pull request where the LLM fixed a real bug in `token_validator.py` (fixed key lookup) but also added redundant `import logging` inside functions and added `logger.debug('Entering function')` everywhere. Clean up the slop relative to `main` without breaking the bug fix. | pass→fail | 5,861 | 5,224 | -11% | 1 | 1 | 0% | 1,018 | 381 | -63% | 0 | 0 | — |
▸case-18 I just finished pruning redundant try-catch blocks and `any` type casts from `service.ts` on my feature branch against `main`. Write the final report following the standard slop cleanup workflow. | fail→fail | 10,484 | 4,896 | -53% | 1 | 1 | 0% | 1,678 | 384 | -77% | 0 | 0 | — |
▸case-19 In `user_service.py`, `validate_user_payload()` is always executed before `create_user()`. An AI assistant added `if user is None: raise ValueError` and `try...except Exception` blocks inside `create_user()`. Clean up this branch against `main`. | fail→fail | 9,555 | 4,872 | -49% | 1 | 1 | 0% | 1,623 | 393 | -76% | 0 | 0 | — |
▸case-20 My branch `feature/auth` is 5 commits behind `main` and has a git merge conflict in `go.mod`. How should I resolve the git rebase conflict using standard git interactive rebase tools? | pass→pass | 10,276 | 7,965 | -22% | 1 | 1 | 0% | 1,937 | 1,376 | -29% | 0 | 0 | — |
▸case-21 Write a comprehensive Jest unit test suite for `auth_controller.ts` covering edge cases like missing authorization headers and expired JWT tokens. | pass→fail | 22,292 | 7,824 | -65% | 1 | 1 | 0% | 5,029 | 545 | -89% | 0 | 0 | — |
▸case-22 We want to migrate our monolithic Express app in `app.js` to a microservice architecture using gRPC. How should we decompose the domain models and route handlers? | pass→fail | 17,838 | 4,446 | -75% | 1 | 1 | 0% | 3,042 | 388 | -87% | 0 | 0 | — |
▸case-23 Our PostgreSQL database queries in `reports.py` are taking 15 seconds to run. How can we analyze and optimize the SQL query execution plan using `EXPLAIN ANALYZE`? | pass→fail | 28,559 | 8,510 | -70% | 1 | 1 | 0% | 2,991 | 438 | -85% | 0 | 0 | — |