# facebookresearch/swe-sweep

> A benchmark for evaluating LLM agents on real repositories: the agent must find and fix as many bugs as possible, with no hints about bug type or location. A thin wrapper around Harbor with final-score computation.

- Magnitude: 0.8 out of 10 — Steady
- Stars: 57 total · +0 stars measured 2026-10-07, 00:20–03:55 UTC, ≈ 1 by evening
- Star trust: star growth looks organic
- Category: Language models · Language: Python · License: MIT · Created: 2026-10-01 · Last push: 2026-10-01
- GitHub: https://github.com/facebookresearch/swe-sweep · Homepage: https://swesweep.com · Page: https://gitnova.dev/en/r/facebookresearch/swe-sweep

## Useful for

- Run an agent on a task under tasks/ and get a bug-fix score
- Compare patches from different models via sweep info --per-task
- Verify the environment and dependencies with sweep infra doctor

## Why it’s here

- Star-counter measurements on 2026-10-07 (UTC), 00:20–03:55: 57 → 57 stars (+0). This is the change over that interval.
- The repository is 6 days old and already has 57 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.
- Hacker News: “Show HN: Benchmark: AI doesn't find bugs unless you tell it what's wrong” — 5 points, 5 days ago.
- Recent forks include notable developers: @mbrukman (394 followers).

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 5
- Issues and pull requests: 2
- Watchers: 0
- Average over the last week: 8 per day
- Usual pace: too little history (under two weeks)
- Stars in the last hour (measured): 0

## Stars per day, last 11 days (oldest → newest, today is partial)

2026-09-27 … 2026-10-07: 0, 0, 0, 1, 22, 21, 7, 1, 4, 1, 0

## Hacker News

- Show HN: Benchmark: AI doesn't find bugs unless you tell it what's wrong — 5 points, 2 comments: https://news.ycombinator.com/item?id=49923102

## More in this category

1. **Niko1221/Strata** — 6.4 · Peaking · Language models · C++ · +221 stars measured 2026-10-07, 00:19–03:47 UTC, ≈ 1,463 by evening
   A C++ local inference engine that runs the 125B MoE model Qwen3.8-Flash-Next on a regular PC with one NVIDIA GPU (12-24 GB) and 64 GB RAM, exposing an OpenAI/Anthropic-compatible API on localhost.
   Full card: https://gitnova.dev/en/r/Niko1221/Strata.md
2. **deepseek-ai/DeepGEMM** — 6.2 · Breakout · Language models · Cuda · +59 stars measured 2026-10-07, 00:20–03:47 UTC, ≈ 275 by evening
   CUDA tensor core kernel library: GEMM (FP8, FP4, BF16), fused MoE, MQA scoring for the indexer. Kernels compile at runtime via DeepJIT, no CUDA build at install time.
   Full card: https://gitnova.dev/en/r/deepseek-ai/DeepGEMM.md
3. **BudEcosystem/Bud-Decision-Studio** — 4.5 · Early signal · Language models · HTML · +0 stars measured 2026-10-07, 00:29–03:48 UTC, ≈ 65 by evening
   Cross-platform desktop app and local server for running decision models: give a situation and typed questions, get calibrated probabilities for each answer in milliseconds.
   Full card: https://gitnova.dev/en/r/BudEcosystem/Bud-Decision-Studio.md
4. **StayLameBro/backburner** — 4.4 · Early signal · Language models · Python · +12 stars measured 2026-10-07, 00:20–03:48 UTC, ≈ 85 by evening
   A llama.cpp fork that plugs an iPhone into a Mac over USB-C and splits the work: the Mac runs layers 1-40, the iPhone runs 41-64 on its GPU, speeding up prefill and allowing context up to 196k-229k tokens. Aimed at people running…
   Full card: https://gitnova.dev/en/r/StayLameBro/backburner.md
5. **Hiteater-wzm/eeo** — 4.4 · Early signal · Language models · JavaScript · +2 stars measured 2026-10-07, 00:20–03:48 UTC, ≈ 33 by evening
   An open community and toolkit for "Everything Engine Optimization": brands publish machine-readable cards about themselves, and the tooling measures how visible a brand is in AI engine answers.
   Full card: https://gitnova.dev/en/r/Hiteater-wzm/eeo.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-07 03:58 UTC, updated every 30 minutes.