WeaveBench/pyproject.toml at main · weavebench/WeaveBench · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
[build-system]
requires = ["setuptools>=68", "wheel"]
build-backend = "setuptools.build_meta"

[project]
name = "weavebench"
version = "0.1.0"
description = "WeaveBench: long-horizon hybrid-interface benchmark (GUI+CLI+code) for computer-use agents"
readme = "README.md"
license = { file = "LICENSE" }
authors = [{ name = "Wanli Li" }]
keywords = ["benchmark", "computer-use-agent", "gui", "cli", "openrouter"]
classifiers = [
  "Intended Audience :: Science/Research",
  "License :: OSI Approved :: MIT License",
  "Programming Language :: Python :: 3",
  "Programming Language :: Python :: 3.10",
  "Programming Language :: Python :: 3.11",
  "Programming Language :: Python :: 3.12",
  "Topic :: Scientific/Engineering :: Artificial Intelligence",
]
requires-python = ">=3.10"
dependencies = [
  # LLM client — only one. All 4 harnesses route through OpenRouter via the
  # OpenAI SDK; Claude / Gemini / GPT etc. are model ids on OpenRouter.
  "openai>=1.40",
  # Dataset / asset download
  "huggingface_hub>=0.24",
  "tqdm>=4.66",
  "requests>=2.31",
  # Bundled desktop_env (OSWorld-derived KVM/Docker provider + Flask shim)
  "gymnasium>=0.29",
  "docker>=7.0",
  "psutil>=5.9",
  "filelock>=3.12",
  "lxml>=4.9",
  "flask>=3.0",
]

[project.optional-dependencies]
osworld-evaluators = [
  # Only needed if you also pip-install upstream OSWorld and use its evaluator
  # pipeline. WeaveBench's own trajectory-aware judge does not need any of these.
  "selenium>=4.20",
  "openpyxl>=3.1",
  "python-docx>=1.0",
  "borb>=2.1",
  "odfpy>=1.4",
  "fastdtw>=0.3",
  "mutagen>=1.47",
  "beautifulsoup4>=4.12",
  "pillow>=10.0",
]
dev = [
  "pytest>=8.0",
  "ruff>=0.4",
]

[project.scripts]
weavebench-run = "weavebench.eval.run_weavebench:main"
weavebench-quickstart = "weavebench.scripts.quickstart:main"
weavebench-download-assets = "weavebench.scripts.download_assets:main"
weavebench-download-dataset = "weavebench.scripts.download_dataset:main"
weavebench-download-vm = "weavebench.scripts.download_vm:main"
weavebench-download-judge = "weavebench.scripts.download_judge:main"
weavebench-demo = "weavebench.scripts.demo:main"

[project.urls]
Homepage = "https://weavebench.github.io"
Repository = "https://github.com/weavebench/WeaveBench"
Dataset = "https://huggingface.co/datasets/wanlilll/WeaveBench"
"Bug Tracker" = "https://github.com/weavebench/WeaveBench/issues"

[tool.setuptools.packages.find]
where = ["."]
include = ["weavebench*"]
exclude = ["tests*", "examples*", "docs*", "scripts*"]

[tool.setuptools.package-data]
"weavebench.assets" = [
  "computer_tool_plugin/*",
  "openclaw_patches/*",
  "codex_patches/*",
  "claudecode_patches/*",
  "hermes_patches/*",
  "weavebench_computer_mcp/*",
]
"weavebench.eval.agent_judge" = ["*.txt"]
"weavebench.desktop_env" = ["*.md"]
"weavebench.desktop_env.providers.docker_lite" = ["*.py", "Dockerfile*"]