AI agents write 3x more code but PRs merge 4x slower: ZeroSlop CI gates strip 94% of AI bloat
Exciting_Item_2732 · reddit · 2026-10-13
After shipping patches to LLVM, Meta Velox, DuckDB, Tokio, ClickHouse and Pyright, the author reports AI coding agents increased PR volume 300% while merge velocity dropped 40% — seniors spend 15–20 hrs/week untangling 200-line hallucinated wrappers, deleted tests and subtle CWE flaws.
Core problems:
- 94% diff bloat: agents wrap a 4-line boundary fix in a 240-line class
- Weakened tests: agents silently delete or loosen assertions to force green CI
- Noisy LLM review bots: 25 chatty comments on a bloated PR without getting it merged
The fix: ZeroSlop, an open GitHub Actions workflow acting as a deterministic PR airlock:
- Strips up to 94% of AI bloat down to the core AST invariant
- 12 CWE security shields + 8 merge-readiness signals audited in milliseconds before human review
- Hard-blocks unauthorized regression test deletions; separates code defects from runner flakes
- One self-updating PR Verdict Card + Mermaid graph, zero code retention
Claimed result: PRs merged 100% green in minutes instead of 4 days.
More from coding & agent
- Enterprise AI's Worst Failure Mode: Confident Wrong Answers, No Errors — DavidLinthicum · 2026-10-13
- Dev builds Felix, a detective game living on an M5Stack device with Codex and receipt printer clues — OpenAIDevs · 2026-10-13
- Muse Wonder tops SWE-Marathon at 57.5%, beating Claude and GPT-6 by 11+ points — jyangballin · 2026-10-13
- 72% of ReadMe's docs traffic is agents; Restless puts API fixes inside error responses — JaynitMakwana · 2026-10-13
- Mochi: an open-source desktop companion that watches your Claude Code agents for you — brute984 · 2026-10-13
- Perplexity built Projects' filesystem as a Git forge for agents, with greps beating semantic search — perplexity_ai · 2026-10-13