README.md

---
pretty_name: AsynCodeBench
license: other
license_name: asyncodebench-apache-2.0-with-upstream-licenses
license_link: https://github.com/KaituoZhang/AsynCodeBench/blob/2a639b049f249e79eb2031876bc06a110be9ac3a/THIRD_PARTY_NOTICES.md
language:
- en
task_categories:
- text-generation
tags:
- code
- software-engineering
- benchmark
- multi-agent
size_categories:
- n<1K
configs:
- config_name: tasks
  default: true
  data_files:
  - split: test
    path: data/tasks-test.parquet
- config_name: scenarios
  data_files:
  - split: test
    path: data/scenarios-test.parquet
- config_name: dependencies
  data_files:
  - split: test
    path: data/dependencies-test.parquet
---

# AsynCodeBench

AsynCodeBench evaluates coding agents under five execution protocols while holding the task set and evaluation contracts fixed. The stable **v0.4.2** release contains **19 coding tasks**. Start with the `tasks` configuration; each row is one real benchmark task.

[**Browse the 19 task examples in Dataset Viewer →**](https://huggingface.co/datasets/PhilipZhang/AsynCodeBench/viewer/tasks/test)

The executable harness, manifests, container references, evaluation commands, and Quick Start are versioned in the [official GitHub repository](https://github.com/KaituoZhang/AsynCodeBench/tree/2a639b049f249e79eb2031876bc06a110be9ac3a). Source repositories are materialized at each row's `base_revision` inside the recorded container image; the Parquet files therefore do not duplicate complete source checkouts or container layers.

## Start here: one task example

The first `tasks` row is `asyncodebench:cachetools`. It contains the full coding prompt in `problem_statement`, the immutable source revision, evaluator command, test targets, and pinned container image/digest.

```python
from datasets import load_dataset

tasks = load_dataset(
    "PhilipZhang/AsynCodeBench",
    "tasks",
    split="test",
    revision="v0.4.2",
)
example = tasks[0]
print(example["task_id"])             # asyncodebench:cachetools
print(example["problem_statement"])   # complete task prompt
print(example["container_image"])     # docker.io/wentingzhao/cachetools:v0
print(example["dependency_count"])    # 5
```

## What the three configurations mean

| Configuration | Rows | Purpose |
|---|---:|---|
| **`tasks`** | **19** | **Primary dataset. One row per coding task: prompt, source revision, tests, evaluator, and container. Most users only need this.** |
| `scenarios` | 95 | Auxiliary protocol table: 19 tasks × 5 execution protocols. These are run conditions, not 95 additional coding tasks. |
| `dependencies` | 52 | Auxiliary annotation table: producer-consumer contracts inside the 19 tasks, with probe tests for ADPR, DRS, CAIL, and SAD metrics. These are not additional tasks. |

All configurations use the `test` split because every row belongs to the evaluation benchmark.

- [Browse task examples](https://huggingface.co/datasets/PhilipZhang/AsynCodeBench/viewer/tasks/test)
- [Browse protocol conditions](https://huggingface.co/datasets/PhilipZhang/AsynCodeBench/viewer/scenarios/test)
- [Browse dependency annotations](https://huggingface.co/datasets/PhilipZhang/AsynCodeBench/viewer/dependencies/test)

## Load auxiliary analysis tables

You do not need these tables to enumerate the 19 coding tasks. Load them when comparing protocols or computing dependency-aware metrics:

```python
scenarios = load_dataset("PhilipZhang/AsynCodeBench", "scenarios", split="test", revision="v0.4.2")
dependencies = load_dataset("PhilipZhang/AsynCodeBench", "dependencies", split="test", revision="v0.4.2")
```

## Run the benchmark

Follow the [GitHub Quick Start](https://github.com/KaituoZhang/AsynCodeBench/blob/2a639b049f249e79eb2031876bc06a110be9ac3a/docs/QUICKSTART.md) for environment setup, model serving, official protocol commands, and result validation. Use the exact container image and digest recorded in the `tasks` configuration.

The five protocols are `single`, `serial_specialists`, `async_private`, `caid_manager` (displayed as Async-RO-Manager), and `async_manager`.

## Version and scope

- Dataset release: `v0.4.2`
- GitHub revision: `2a639b049f249e79eb2031876bc06a110be9ac3a`
- Release status: stable
- Official task count: 19
- Official protocol conditions: 95
- Official dependency points: 52

Candidate manifests outside the v0.4 release index are excluded. In particular, `apache-tvm-20409` is not part of this dataset release.

## License

The original AsynCodeBench material is Apache-2.0. Referenced upstream tasks and repositories retain their own licenses. See [LICENSE](https://github.com/KaituoZhang/AsynCodeBench/blob/2a639b049f249e79eb2031876bc06a110be9ac3a/LICENSE) and [THIRD_PARTY_NOTICES.md](https://github.com/KaituoZhang/AsynCodeBench/blob/2a639b049f249e79eb2031876bc06a110be9ac3a/THIRD_PARTY_NOTICES.md).

## Citation

If you use AsynCodeBench, cite the software release using the repository's [CITATION.cff](https://github.com/KaituoZhang/AsynCodeBench/blob/2a639b049f249e79eb2031876bc06a110be9ac3a/CITATION.cff).