{
  "added_at": "2026-11-02",
  "additional_evidence_urls": [
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/README.md#L124-L126",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/CHANGELOG.md",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/setup.py",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/pyproject.toml",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/refinery/__init__.py",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/.github/workflows/test.yml",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/samples/__init__.py",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/refinery/lib/thirdparty/__init__.py",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/refinery/lib/seven/__init__.py",
    "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/LICENSE.md",
    "https://pypi.org/project/binary-refinery/0.11.2/",
    "https://github.com/binref/agent/blob/be019a500f1827c9359c51e9879d6144cb7453de/README.md"
  ],
  "category": "native-analysis",
  "description_en": "Large collection of composable Python command-line units for transforming binary data: encodings, compression, encryption, archive and executable-format extraction, carving and pattern search. Each unit reads stdin and writes stdout, so units chain into shell pipelines, while framing, multibin arguments and meta variables let one pipeline process many chunks, and the binref command searches unit help. The author describes it as a command-line take on CyberChef focused on malware triage; the README also points to a separate binref/agent repository of instructions for language-model agents that use refinery.",
  "description_zh": "由大量可组合的 Python 命令行单元组成的二进制数据转换工具集，涵盖编码、压缩、加密、归档与可执行文件格式提取、数据切割（carving）和模式搜索。各单元从标准输入读取、向标准输出写入，因此可串联成 shell 管道；分帧（framing）、multibin 参数和元变量使一条管道能处理多个数据块，binref 命令可检索各单元的帮助文本。作者称其为面向恶意软件初步分析、在命令行上实现类似 CyberChef 功能的工具；README 还指向独立的 binref/agent 仓库，其中提供供语言模型智能体使用 refinery 的说明。",
  "entry_url": "https://awesomerea.org/resources/binary-refinery/",
  "entry_url_zh": "https://awesomerea.org/zh/resources/binary-refinery/",
  "evidence_url": "https://github.com/binref/refinery/blob/13d8c68863c3d11d4e6cd36007fdd6c53459eda6/README.md",
  "group": "tools",
  "id": "binary-refinery",
  "json_url": "https://awesomerea.org/resources/binary-refinery/record.json",
  "kind": "tool",
  "markdown_url": "https://awesomerea.org/resources/binary-refinery/index.md",
  "markdown_url_zh": "https://awesomerea.org/zh/resources/binary-refinery/index.md",
  "name": "Binary Refinery",
  "name_zh": "Binary Refinery 二进制数据转换工具集",
  "record_type": "resource",
  "relationship": "foundation",
  "requirements": [
    "Latest release at review: 0.11.2, uploaded to PyPI on 2026-07-14 and requiring Python 3.10+. The reviewed master commit of 2026-10-11 is 1,138 commits past that tag and still reports version 0.11.2, while its changelog already lists breaking 0.11.3 changes, including Python 3.11+ and renamed meta variables. The README says refinery is still in alpha and that units and parameters can be renamed or removed.",
    "Installed with pip as binary-refinery, preferably into its own virtual environment. The default, extended and all extras add progressively more optional dependencies, and units name a missing dependency when run. Installation creates one shell command per unit, optionally with a prefix to avoid name clashes. The README rates Bash and CMD support as good, PowerShell 7.4+ as reasonable, and Zsh and Fish as having minor issues. Several decompression and cipher routines have Cython extensions next to pure-Python versions.",
    "Inputs are the analyst's own files or data piped on stdin. CI runs pytest on Windows, macOS and Ubuntu with Python 3.11 and 3.13. Tests that need real-world inputs fetch encrypted files from the separate binref/refinery-test-data repository. Unit output depends on the parameters chosen and needs analyst validation.",
    "LICENSE.md is the 3-Clause BSD License, which GitHub does not classify. refinery/lib/thirdparty keeps third-party modules under their own licenses, and refinery/lib/seven ports compression code from 7-Zip."
  ],
  "requirements_zh": [
    "审阅时最新版本为 0.11.2，于 2026-07-14 上传到 PyPI，要求 Python 3.10+。所审阅的 master 分支提交日期为 2026-10-11，比该标签多 1,138 个提交，版本号仍为 0.11.2，但其更新日志已列出 0.11.3 的不兼容变更，包括要求 Python 3.11+ 以及元变量改名。README 称 refinery 仍处于 alpha 阶段，单元和参数可能被改名或移除。",
    "通过 pip 安装 binary-refinery，建议装在独立的虚拟环境中。default、extended 和 all 三组附加选项依次加入更多可选依赖，运行单元时会提示缺少哪个依赖。安装时会为每个单元创建一条 shell 命令，可加前缀以避免命名冲突。README 将 Bash 和 CMD 的支持评为良好，PowerShell 7.4+ 为尚可，Zsh 和 Fish 存在小问题。若干解压缩和加密例程同时提供 Cython 扩展和纯 Python 版本。",
    "输入为分析者自有的文件或经标准输入管道传入的数据。CI 在 Windows、macOS 和 Ubuntu 上以 Python 3.11 和 3.13 运行 pytest。需要真实样本输入的测试会从独立的 binref/refinery-test-data 仓库获取加密文件。单元输出取决于所选参数，需要分析者核实。",
    "LICENSE.md 为三条款 BSD 许可证，GitHub 未能识别。refinery/lib/thirdparty 中的第三方模块保留各自的许可证，refinery/lib/seven 移植了 7-Zip 的压缩代码。"
  ],
  "targets": [
    "binary"
  ],
  "url": "https://github.com/binref/refinery",
  "verification": "source-reviewed",
  "verified_at": "2026-10-20"
}