├── .difyignore ├── .env.example ├── .gitignore ├── GUIDE.md ├── LICENSE ├── PRIVACY.md ├── README.md ├── _assets └── icon.svg ├── img ├── actions.png ├── retrieve.png └── upload.png ├── knowledge-manager.difypkg ├── knowledge-upload.difypkg ├── main.py ├── manifest.yaml ├── plugin.yaml ├── provider ├── knowledge.py └── knowledge.yaml ├── requirements.txt ├── test_tool.py ├── test_tool.py.combined ├── test_tool.py.retrieve ├── test_tool.py.upload └── tools ├── knowledge_retrieve.py ├── knowledge_retrieve.yaml ├── knowledge_upload.py └── knowledge_upload.yaml /.difyignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # Distribution / packaging 7 | .Python 8 | build/ 9 | develop-eggs/ 10 | dist/ 11 | downloads/ 12 | eggs/ 13 | .eggs/ 14 | lib/ 15 | lib64/ 16 | parts/ 17 | sdist/ 18 | var/ 19 | wheels/ 20 | share/python-wheels/ 21 | *.egg-info/ 22 | .installed.cfg 23 | *.egg 24 | MANIFEST 25 | 26 | # PyInstaller 27 | # Usually these files are written by a python script from a template 28 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 29 | *.manifest 30 | *.spec 31 | 32 | # Installer logs 33 | pip-log.txt 34 | pip-delete-this-directory.txt 35 | 36 | # Unit test / coverage reports 37 | htmlcov/ 38 | .tox/ 39 | .nox/ 40 | .coverage 41 | .coverage.* 42 | .cache 43 | nosetests.xml 44 | coverage.xml 45 | *.cover 46 | *.py,cover 47 | .hypothesis/ 48 | .pytest_cache/ 49 | cover/ 50 | 51 | # Translations 52 | *.mo 53 | *.pot 54 | 55 | # Django stuff: 56 | *.log 57 | local_settings.py 58 | db.sqlite3 59 | db.sqlite3-journal 60 | 61 | # Flask stuff: 62 | instance/ 63 | .webassets-cache 64 | 65 | # Scrapy stuff: 66 | .scrapy 67 | 68 | # Sphinx documentation 69 | docs/_build/ 70 | 71 | # PyBuilder 72 | .pybuilder/ 73 | target/ 74 | 75 | # Jupyter Notebook 76 | .ipynb_checkpoints 77 | 78 | # IPython 79 | profile_default/ 80 | ipython_config.py 81 | 82 | # pyenv 83 | # For a library or package, you might want to ignore these files since the code is 84 | # intended to run in multiple environments; otherwise, check them in: 85 | .python-version 86 | 87 | # pipenv 88 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 89 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 90 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 91 | # install all needed dependencies. 92 | Pipfile.lock 93 | 94 | # UV 95 | # Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control. 96 | # This is especially recommended for binary packages to ensure reproducibility, and is more 97 | # commonly ignored for libraries. 98 | uv.lock 99 | 100 | # poetry 101 | # Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control. 102 | # This is especially recommended for binary packages to ensure reproducibility, and is more 103 | # commonly ignored for libraries. 104 | # https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control 105 | poetry.lock 106 | 107 | # pdm 108 | # Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control. 109 | #pdm.lock 110 | # pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it 111 | # in version control. 112 | # https://pdm.fming.dev/latest/usage/project/#working-with-version-control 113 | .pdm.toml 114 | .pdm-python 115 | .pdm-build/ 116 | 117 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm 118 | __pypackages__/ 119 | 120 | # Celery stuff 121 | celerybeat-schedule 122 | celerybeat.pid 123 | 124 | # SageMath parsed files 125 | *.sage.py 126 | 127 | # Environments 128 | .env 129 | .venv 130 | env/ 131 | venv/ 132 | ENV/ 133 | env.bak/ 134 | venv.bak/ 135 | 136 | # Spyder project settings 137 | .spyderproject 138 | .spyproject 139 | 140 | # Rope project settings 141 | .ropeproject 142 | 143 | # mkdocs documentation 144 | /site 145 | 146 | # mypy 147 | .mypy_cache/ 148 | .dmypy.json 149 | dmypy.json 150 | 151 | # Pyre type checker 152 | .pyre/ 153 | 154 | # pytype static type analyzer 155 | .pytype/ 156 | 157 | # Cython debug symbols 158 | cython_debug/ 159 | 160 | # PyCharm 161 | # JetBrains specific template is maintained in a separate JetBrains.gitignore that can 162 | # be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore 163 | # and can be added to the global gitignore or merged into this file. For a more nuclear 164 | # option (not recommended) you can uncomment the following to ignore the entire idea folder. 165 | .idea/ 166 | 167 | # Vscode 168 | .vscode/ 169 | 170 | # Git 171 | .git/ 172 | .gitignore 173 | 174 | # Mac 175 | .DS_Store 176 | 177 | # Windows 178 | Thumbs.db 179 | -------------------------------------------------------------------------------- /.env.example: -------------------------------------------------------------------------------- 1 | INSTALL_METHOD=remote 2 | REMOTE_INSTALL_HOST=debug.dify.ai 3 | REMOTE_INSTALL_PORT=5003 4 | REMOTE_INSTALL_KEY=********-****-****-****-************ 5 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | downloads/ 15 | eggs/ 16 | .eggs/ 17 | lib/ 18 | lib64/ 19 | parts/ 20 | sdist/ 21 | var/ 22 | wheels/ 23 | share/python-wheels/ 24 | *.egg-info/ 25 | .installed.cfg 26 | *.egg 27 | MANIFEST 28 | 29 | # PyInstaller 30 | # Usually these files are written by a python script from a template 31 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 32 | *.manifest 33 | *.spec 34 | 35 | # Installer logs 36 | pip-log.txt 37 | pip-delete-this-directory.txt 38 | 39 | # Unit test / coverage reports 40 | htmlcov/ 41 | .tox/ 42 | .nox/ 43 | .coverage 44 | .coverage.* 45 | .cache 46 | nosetests.xml 47 | coverage.xml 48 | *.cover 49 | *.py,cover 50 | .hypothesis/ 51 | .pytest_cache/ 52 | cover/ 53 | 54 | # Translations 55 | *.mo 56 | *.pot 57 | 58 | # Django stuff: 59 | *.log 60 | local_settings.py 61 | db.sqlite3 62 | db.sqlite3-journal 63 | 64 | # Flask stuff: 65 | instance/ 66 | .webassets-cache 67 | 68 | # Scrapy stuff: 69 | .scrapy 70 | 71 | # Sphinx documentation 72 | docs/_build/ 73 | 74 | # PyBuilder 75 | .pybuilder/ 76 | target/ 77 | 78 | # Jupyter Notebook 79 | .ipynb_checkpoints 80 | 81 | # IPython 82 | profile_default/ 83 | ipython_config.py 84 | 85 | # pyenv 86 | # For a library or package, you might want to ignore these files since the code is 87 | # intended to run in multiple environments; otherwise, check them in: 88 | # .python-version 89 | 90 | # pipenv 91 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 92 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 93 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 94 | # install all needed dependencies. 95 | #Pipfile.lock 96 | 97 | # UV 98 | # Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control. 99 | # This is especially recommended for binary packages to ensure reproducibility, and is more 100 | # commonly ignored for libraries. 101 | #uv.lock 102 | 103 | # poetry 104 | # Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control. 105 | # This is especially recommended for binary packages to ensure reproducibility, and is more 106 | # commonly ignored for libraries. 107 | # https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control 108 | #poetry.lock 109 | 110 | # pdm 111 | # Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control. 112 | #pdm.lock 113 | # pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it 114 | # in version control. 115 | # https://pdm.fming.dev/latest/usage/project/#working-with-version-control 116 | .pdm.toml 117 | .pdm-python 118 | .pdm-build/ 119 | 120 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm 121 | __pypackages__/ 122 | 123 | # Celery stuff 124 | celerybeat-schedule 125 | celerybeat.pid 126 | 127 | # SageMath parsed files 128 | *.sage.py 129 | 130 | # Environments 131 | .env 132 | .venv 133 | env/ 134 | venv/ 135 | ENV/ 136 | env.bak/ 137 | venv.bak/ 138 | 139 | # Spyder project settings 140 | .spyderproject 141 | .spyproject 142 | 143 | # Rope project settings 144 | .ropeproject 145 | 146 | # mkdocs documentation 147 | /site 148 | 149 | # mypy 150 | .mypy_cache/ 151 | .dmypy.json 152 | dmypy.json 153 | 154 | # Pyre type checker 155 | .pyre/ 156 | 157 | # pytype static type analyzer 158 | .pytype/ 159 | 160 | # Cython debug symbols 161 | cython_debug/ 162 | 163 | # PyCharm 164 | # JetBrains specific template is maintained in a separate JetBrains.gitignore that can 165 | # be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore 166 | # and can be added to the global gitignore or merged into this file. For a more nuclear 167 | # option (not recommended) you can uncomment the following to ignore the entire idea folder. 168 | .idea/ 169 | 170 | # Vscode 171 | .vscode/ 172 | -------------------------------------------------------------------------------- /GUIDE.md: -------------------------------------------------------------------------------- 1 | # Dify知识库管理插件使用指南 2 | 3 | ## 简介 4 | 5 | Dify知识库管理插件是一个用于管理Dify知识库的综合工具。它支持创建知识库、上传文本内容以及从知识库检索信息，提供简单的界面来配置各种操作参数。 6 | 7 | ## 安装前准备 8 | 9 | 在安装和使用此插件前，您需要： 10 | 11 | 1. 拥有Dify平台账号 12 | 2. 获取知识库API Key 13 | 14 | ## 安装步骤 15 | 16 | 1. 在Dify平台中，点击右上角的"插件"按钮 17 | 2. 选择"本地上传"选项 18 | 3. 上传插件文件（.difypkg格式） 19 | 4. 完成安装后，您需要配置API Key才能使用插件 20 | 21 | ## 配置API Key 22 | 23 | 1. 在插件列表中找到"Dify知识库管理"插件 24 | 2. 点击"配置"按钮 25 | 3. 在弹出的窗口中输入您的知识库API Key 26 | 4. 点击"保存"按钮完成配置 27 | 28 | ## 如何获取知识库API Key 29 | 30 | 1. 登录Dify平台 31 | 2. 导航到知识库页面 32 | 3. 从左侧导航切换到**API访问**页面 33 | 4. 在**API Keys**部分管理访问凭证 34 | 5. 如果没有API Key，点击"创建"按钮生成一个新的API Key 35 | 36 | ## 如何获取知识库ID 37 | 38 | 知识库ID可以从知识库URL获取，例如： 39 | `https://app.dify.ai/datasets/12345678-1234-1234-1234-123456789012` 40 | 其中`12345678-1234-1234-1234-123456789012`就是知识库ID。 41 | 42 | ## 使用方法 43 | 44 | ### 上传功能 45 | 46 | #### 在Agent应用中使用 47 | 48 | 1. 创建或编辑一个Agent应用 49 | 2. 在应用编排页面底部找到"工具"选项 50 | 3. 选择"知识库上传"工具 51 | 4. 配置工具参数： 52 | - 知识库名称：要创建的知识库名称 53 | - 描述：知识库的描述（可选） 54 | - 文档名称：要创建的文档名称 55 | - 文本内容：要上传的文本内容 56 | - 权限：选择知识库权限（仅自己或公开可读） 57 | - 索引技术：选择高质量(high_quality)或经济(economy)模式 58 | 5. 保存配置 59 | 60 | #### 在Chatflow/Workflow应用中使用 61 | 62 | 1. 在工作流编排画布中，点击节点末尾的"+"号 63 | 2. 选择"知识库上传"工具 64 | 3. 将其连接到上游节点 65 | 4. 配置工具参数 66 | 5. 保存配置 67 | 68 | ### 检索功能 69 | 70 | #### 在Agent应用中使用 71 | 72 | 1. 创建或编辑一个Agent应用 73 | 2. 在应用编排页面底部找到"工具"选项 74 | 3. 选择"知识库检索"工具 75 | 4. 配置工具参数： 76 | - 知识库ID：要检索的知识库ID 77 | - 查询内容：在知识库中搜索的查询内容 78 | - 搜索方法：选择搜索方法（关键词检索、语义检索、全文检索或混合检索） 79 | - 启用重排序：是否启用搜索结果重排序 80 | - 结果数量：返回的结果数量 81 | - 启用分数阈值：是否启用分数阈值过滤 82 | - 分数阈值：结果的最小分数阈值 83 | 5. 保存配置 84 | 85 | #### 在Chatflow/Workflow应用中使用 86 | 87 | 1. 在工作流编排画布中，点击节点末尾的"+"号 88 | 2. 选择"知识库检索"工具 89 | 3. 将其连接到上游节点 90 | 4. 配置工具参数 91 | 5. 保存配置 92 | 93 | ## 上传功能参数说明 94 | 95 | - **知识库名称**（必填）：要创建的知识库名称 96 | - **描述**（可选）：知识库的描述 97 | - **文档名称**（必填）：要创建的文档名称 98 | - **文本内容**（必填）：要上传的文本内容 99 | - **权限**（必填）：知识库的权限设置，可选值： 100 | - `only_me`：仅自己可见，只有创建者可以访问 101 | - `publicly_readable`：公开可读，所有人可以读取 102 | - **索引技术**（必填）：选择索引技术，可选值： 103 | - `high_quality`：高质量模式，提供更准确的检索结果，但处理时间较长 104 | - `economy`：经济模式，处理速度较快，但检索准确性可能略低 105 | 106 | ## 检索功能参数说明 107 | 108 | - **知识库ID**（必填）：要检索的知识库ID 109 | - **查询内容**（必填）：在知识库中搜索的查询内容 110 | - **搜索方法**（可选，默认为语义检索）：用于搜索知识库的方法，可选值： 111 | - `keyword_search`：关键词检索，基于关键词匹配 112 | - `semantic_search`：语义检索，基于语义理解 113 | - `full_text_search`：全文检索，搜索整个文本内容 114 | - `hybrid_search`：混合检索，结合关键词和语义检索 115 | - **启用重排序**（可选，默认为否）：是否启用搜索结果重排序 116 | - **结果数量**（可选，默认为3）：返回的结果数量 117 | - **启用分数阈值**（可选，默认为否）：是否启用分数阈值过滤 118 | - **分数阈值**（可选，默认为0.5）：结果的最小分数阈值，范围0-1 119 | 120 | ## 上传工作流程 121 | 122 | 1. 创建新的知识库：插件会根据提供的名称和描述创建一个新的知识库 123 | 2. 创建文档：在知识库中创建一个空文档 124 | 3. 上传文本内容：将文本内容上传到创建的文档中 125 | 4. 处理文档：Dify平台会自动处理上传的文本内容，包括文本分段和索引 126 | 5. 返回结果：插件会返回处理状态和结果 127 | 128 | ## 检索工作流程 129 | 130 | 1. 提供知识库ID和查询内容 131 | 2. 选择搜索方法和其他参数 132 | 3. 执行知识库检索 133 | 4. 返回检索结果和相关信息 134 | 135 | ## 搜索方法说明 136 | 137 | - **关键词检索**：基于关键词匹配，适合精确查找特定术语或短语 138 | - **语义检索**：基于语义理解，能够找到语义相关但可能不包含完全相同关键词的内容 139 | - **全文检索**：搜索整个文本内容，适合需要在大量文本中查找信息 140 | - **混合检索**：结合关键词和语义检索，提供更全面的搜索结果 141 | 142 | ## 常见问题 143 | 144 | ### 上传或检索失败，提示"API Key validation failed" 145 | 146 | 请检查您提供的API Key是否正确，以及是否有足够的权限访问知识库。 147 | 148 | ### 上传成功后，文档在知识库中不可见 149 | 150 | 文档上传后需要一段时间进行处理和索引。您可以在知识库页面查看文档的处理状态。 151 | 152 | ### 检索失败，提示"知识库不存在或无权访问" 153 | 154 | 请确保您提供的知识库ID是正确的，并且您的API Key有权限访问该知识库。 155 | 156 | ### 检索结果不符合预期 157 | 158 | 尝试调整搜索方法、启用重排序或调整分数阈值，以获得更符合预期的结果。 159 | 160 | ## 技术支持 161 | 162 | 如果您在使用过程中遇到任何问题，请联系插件作者或在Dify社区论坛中提问。 -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2025 Steven Lynn 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /PRIVACY.md: -------------------------------------------------------------------------------- 1 | # 隐私政策 2 | 3 | ## 数据收集和使用 4 | 5 | 本插件需要访问以下信息： 6 | 7 | 1. **Dify知识库API Key**：用于验证用户身份并授权访问Dify知识库。API Key存储在用户的Dify工作区中，不会被插件保存或传输到其他地方。 8 | 9 | 2. **文本内容**：插件需要读取用户提供的文本内容，以便上传到Dify知识库。文本内容仅用于上传目的，不会被插件保存或用于其他目的。 10 | 11 | 3. **查询内容**：插件需要读取用户提供的查询内容，以便在Dify知识库中进行检索。查询内容仅用于检索目的，不会被插件保存或用于其他目的。 12 | 13 | 4. **知识库ID**：插件需要知道用户想要检索的知识库ID。知识库ID仅用于检索目的，不会被插件保存或用于其他目的。 14 | 15 | ## 数据传输 16 | 17 | 所有数据传输都通过HTTPS协议进行加密。插件直接与Dify API通信，不经过任何第三方服务器。 18 | 19 | ## 数据存储 20 | 21 | 插件本身不存储任何用户数据。所有上传的文本内容都存储在用户的Dify知识库中，受Dify平台的隐私政策保护。所有检索操作都是实时进行的，结果直接返回给用户，不会在插件中保存。 22 | 23 | ## 第三方服务 24 | 25 | 本插件使用Dify API服务。使用本插件即表示您同意Dify的服务条款和隐私政策。 -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Dify Knowledge Base Manager Plugin 2 | 3 | A plugin for managing Dify knowledge base - creating knowledge bases, uploading text content, and retrieving information. 4 | ![](./img/actions.png) 5 | 6 | ## Prerequisites 7 | 8 | Before using this plugin, you need: 9 | 10 | 1. A Dify platform account 11 | 2. Knowledge base API Key 12 | 13 | ## How to Get API Key 14 | 15 | 1. Log in to [Dify.ai](https://cloud.dify.ai) 16 | 2. Navigate to the knowledge base page 17 | 3. Switch to the **API Access** page from the left navigation 18 | 4. Manage access credentials in the **API Keys** section 19 | 24 | 25 | ## How to Get Knowledge Base ID 26 | 27 | The knowledge base ID can be obtained from the knowledge base URL, for example: 28 | `https://app.dify.ai/datasets/12345678-1234-1234-1234-123456789012` 29 | where `12345678-1234-1234-1234-123456789012` is the knowledge base ID. 30 | 31 | ## Upload Tool Parameters 32 | 33 | ![](./img/upload.png) 34 | - **Knowledge Base Name**: The name of the knowledge base to create 35 | - **Description**: Description of the knowledge base (optional) 36 | - **Document Name**: The name of the document to create 37 | - **Text Content**: The text content to upload 38 | - **Permission**: Knowledge base permission settings 39 | - only_me: Only the creator can access 40 | - publicly_readable: Everyone can read 41 | - **Indexing Technology**: Choose high_quality or economy mode 42 | 43 | ## Retrieve Tool Parameters 44 | 45 | ![](./img/retrieve.png) 46 | - **Knowledge Base ID**: The ID of the knowledge base to retrieve from (required) 47 | - **Query**: The query to search for in the knowledge base (required) 48 | - **Search Method**: The method to use for searching the knowledge base (optional, default is semantic search) 49 | - keyword_search: Keyword search, based on keyword matching 50 | - semantic_search: Semantic search, based on semantic understanding 51 | - full_text_search: Full text search, searches the entire text content 52 | - hybrid_search: Hybrid search, combines keyword and semantic search 53 | - **Enable Reranking**: Whether to enable reranking of search results (optional, default is false) 54 | - **Number of Results**: The number of results to return (optional, default is 3) 55 | - **Enable Score Threshold**: Whether to enable score threshold filtering (optional, default is false) 56 | - **Score Threshold**: The minimum score threshold for results (0-1) (optional, default is 0.5) 57 | 58 | ## Upload Output 59 | 60 | The upload tool returns a JSON response with the following structure: 61 | ```json 62 | { 63 | "status": "success", 64 | "knowledge_base_id": "12345678-1234-1234-1234-123456789012", 65 | "knowledge_base": { 66 | "id": "12345678-1234-1234-1234-123456789012", 67 | "name": "Knowledge Base Name" 68 | }, 69 | "document": { 70 | "id": "document-id", 71 | "name": "Document Name", 72 | "batch": "batch-id", 73 | "status": "completed" 74 | } 75 | } 76 | ``` 77 | 78 | ## Retrieve Output 79 | 80 | The retrieve tool returns a JSON response with the following structure: 81 | ```json 82 | { 83 | "status": "success", 84 | "query": "Query content", 85 | "knowledge_base_id": "Knowledge base ID", 86 | "results": [ 87 | { 88 | "segment": { 89 | "id": "Segment ID", 90 | "content": "Segment content", 91 | "document": { 92 | "id": "Document ID", 93 | "name": "Document name" 94 | } 95 | }, 96 | "score": 0.95 97 | } 98 | ] 99 | } 100 | ``` 101 | 102 | The `knowledge_base_id` field can be used for further operations with the knowledge base. 103 | 104 | ## Notes 105 | 106 | - Text content requires some time for processing and indexing after upload 107 | - Processing large amounts of text may take longer 108 | - If processing is not complete, the plugin will return the current status, and you can check the processing results later on the Dify platform 109 | - Text content will be automatically segmented for processing, using automatic mode by default 110 | 111 | ## Supported File Formats 112 | 113 | - Text files (.txt) 114 | - PDF files (.pdf) 115 | - Word documents (.doc, .docx) 116 | - Markdown files (.md, .markdown) 117 | - HTML files (.html, .htm) 118 | - Excel files (.xlsx) 119 | - CSV files (.csv) 120 | 121 | 122 | 123 | -------------------------------------------------------------------------------- /_assets/icon.svg: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | -------------------------------------------------------------------------------- /img/actions.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/stvlynn/better-knowledge-base/d78e166819ff14e774f8e612ca7e4b1255b59563/img/actions.png -------------------------------------------------------------------------------- /img/retrieve.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/stvlynn/better-knowledge-base/d78e166819ff14e774f8e612ca7e4b1255b59563/img/retrieve.png -------------------------------------------------------------------------------- /img/upload.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/stvlynn/better-knowledge-base/d78e166819ff14e774f8e612ca7e4b1255b59563/img/upload.png -------------------------------------------------------------------------------- /knowledge-manager.difypkg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/stvlynn/better-knowledge-base/d78e166819ff14e774f8e612ca7e4b1255b59563/knowledge-manager.difypkg -------------------------------------------------------------------------------- /knowledge-upload.difypkg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/stvlynn/better-knowledge-base/d78e166819ff14e774f8e612ca7e4b1255b59563/knowledge-upload.difypkg -------------------------------------------------------------------------------- /main.py: -------------------------------------------------------------------------------- 1 | from dify_plugin import Plugin, DifyPluginEnv 2 | 3 | plugin = Plugin(DifyPluginEnv(MAX_REQUEST_TIMEOUT=120)) 4 | 5 | if __name__ == '__main__': 6 | plugin.run() 7 | -------------------------------------------------------------------------------- /manifest.yaml: -------------------------------------------------------------------------------- 1 | version: 0.0.1 2 | type: plugin 3 | author: stvlynn 4 | name: knowledge-manager 5 | label: 6 | en_US: Dify Knowledge Base Manager 7 | ja_JP: Dify Knowledge Base Manager 8 | zh_Hans: Dify知识库管理器 9 | pt_BR: Dify Knowledge Base Manager 10 | description: 11 | en_US: A tool to manage Dify Knowledge Base - create, upload text content, and retrieve information. 12 | ja_JP: A tool to manage Dify Knowledge Base - create, upload text content, and retrieve information. 13 | zh_Hans: 一个管理Dify知识库的工具 - 创建、上传文本内容和检索信息。 14 | pt_BR: A tool to manage Dify Knowledge Base - create, upload text content, and retrieve information. 15 | icon: icon.svg 16 | resource: 17 | memory: 268435456 18 | permission: {} 19 | plugins: 20 | tools: 21 | - provider/knowledge.yaml 22 | meta: 23 | version: 0.0.1 24 | arch: 25 | - amd64 26 | - arm64 27 | runner: 28 | language: python 29 | version: "3.12" 30 | entrypoint: main 31 | created_at: 2025-03-04T23:40:41.394653+08:00 32 | privacy: PRIVACY.md 33 | verified: false 34 | -------------------------------------------------------------------------------- /plugin.yaml: -------------------------------------------------------------------------------- 1 | name: knowledge-manager 2 | version: 0.0.1 3 | description: 4 | en_US: Manage Dify knowledge base - create, upload text content, and retrieve information 5 | zh_Hans: 管理Dify知识库 - 创建、上传文本内容和检索信息 6 | author: stvlynn 7 | avatar: ./_assets/icon.svg 8 | category: knowledge 9 | type: tool 10 | env: 11 | - name: DIFY_KNOWLEDGE_API_KEY 12 | description: 13 | en_US: Dify Knowledge API Key 14 | zh_Hans: Dify知识库API密钥 15 | type: secret 16 | required: true 17 | help: 18 | en_US: You can get the API Key from the API Access page in the Dify knowledge base 19 | zh_Hans: 您可以从Dify知识库的API访问页面获取API密钥 -------------------------------------------------------------------------------- /provider/knowledge.py: -------------------------------------------------------------------------------- 1 | from typing import Any 2 | import requests 3 | import os 4 | 5 | from dify_plugin import ToolProvider 6 | from dify_plugin.errors.tool import ToolProviderCredentialValidationError 7 | 8 | 9 | class KnowledgeProvider(ToolProvider): 10 | def _validate_credentials(self, credentials: dict[str, Any]) -> None: 11 | try: 12 | api_key = credentials.get('api_key') 13 | if not api_key: 14 | raise ValueError("API Key is required") 15 | 16 | # 将API Key保存到环境变量中，以便工具可以访问 17 | os.environ['DIFY_KNOWLEDGE_API_KEY'] = api_key 18 | 19 | # 尝试使用API Key获取知识库列表，验证API Key是否有效 20 | headers = { 21 | 'Authorization': f'Bearer {api_key}', 22 | 'Content-Type': 'application/json' 23 | } 24 | 25 | response = requests.get('https://api.dify.ai/v1/datasets?page=1&limit=1', headers=headers) 26 | 27 | if response.status_code != 200: 28 | error_data = response.json() 29 | raise ValueError(f"API Key validation failed: {error_data.get('message', 'Unknown error')}") 30 | 31 | except Exception as e: 32 | raise ToolProviderCredentialValidationError(str(e)) 33 | -------------------------------------------------------------------------------- /provider/knowledge.yaml: -------------------------------------------------------------------------------- 1 | identity: 2 | author: stvlynn 3 | name: knowledge 4 | label: 5 | en_US: Dify Knowledge Base 6 | zh_Hans: Dify 知识库 7 | description: 8 | en_US: A tool to manage Dify Knowledge Base - upload content and retrieve information 9 | zh_Hans: 一个管理Dify知识库的工具 - 上传内容和检索信息 10 | icon: icon.svg 11 | tags: 12 | - productivity 13 | - education 14 | credentials_for_provider: 15 | api_key: 16 | type: secret-input 17 | required: true 18 | label: 19 | en_US: Dify Knowledge Base API Key 20 | zh_Hans: Dify 知识库 API Key 21 | placeholder: 22 | en_US: Please input your Dify Knowledge Base API Key 23 | zh_Hans: 请输入您的Dify知识库API Key 24 | help: 25 | en_US: Get your API Key from Dify Knowledge Base API Access page 26 | zh_Hans: 从Dify知识库API访问页面获取您的API Key 27 | tools: 28 | - tools/knowledge_upload.yaml 29 | - tools/knowledge_retrieve.yaml 30 | extra: 31 | python: 32 | source: provider/knowledge.py 33 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | dify-plugin==0.0.1b70 2 | requests>=2.31.0 3 | -------------------------------------------------------------------------------- /test_tool.py: -------------------------------------------------------------------------------- 1 | import os 2 | import json 3 | import time 4 | import requests 5 | 6 | # Test parameters 7 | knowledge_base_name = "Test KB" 8 | description = "" 9 | document_name = "Test Doc" 10 | text_content = "This is a test content for the document." 11 | permission = "only_me" 12 | indexing_technique = "high_quality" 13 | 14 | # Set API Key 15 | api_key = "test_key" # Replace with your actual API Key 16 | 17 | # Set request headers 18 | headers = { 19 | 'Authorization': f'Bearer {api_key}', 20 | 'Content-Type': 'application/json' 21 | } 22 | 23 | print("Starting API request test...") 24 | 25 | # Step 1: Create knowledge base 26 | print("Step 1: Creating knowledge base") 27 | url = "https://api.dify.ai/v1/datasets" 28 | payload = { 29 | "name": knowledge_base_name, 30 | "description": description, 31 | "permission": permission, 32 | "indexing_technique": indexing_technique, 33 | "provider": "vendor" 34 | } 35 | 36 | print(f"Request URL: {url}") 37 | print(f"Request parameters: {payload}") 38 | 39 | try: 40 | response = requests.post(url, headers=headers, json=payload) 41 | print(f"Response status code: {response.status_code}") 42 | print(f"Response content: {response.text}") 43 | 44 | if response.status_code == 200: 45 | result = response.json() 46 | dataset_id = result.get('id') 47 | print(f"Knowledge base created successfully, ID: {dataset_id}") 48 | 49 | # Step 2: Create document by text 50 | print("\nStep 2: Creating document") 51 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/document/create-by-text" 52 | 53 | # Prepare processing rules 54 | process_rule = { 55 | "mode": "automatic" 56 | } 57 | 58 | payload = { 59 | "name": document_name, 60 | "text": text_content, 61 | "indexing_technique": indexing_technique, 62 | "process_rule": process_rule 63 | } 64 | 65 | print(f"Request URL: {url}") 66 | print(f"Request parameters: {payload}") 67 | 68 | response = requests.post(url, headers=headers, json=payload) 69 | print(f"Response status code: {response.status_code}") 70 | print(f"Response content: {response.text}") 71 | 72 | if response.status_code == 200: 73 | result = response.json() 74 | document = result.get('document', {}) 75 | document_id = document.get('id') 76 | batch = result.get('batch', '') 77 | print(f"Document created successfully, ID: {document_id}, Batch: {batch}") 78 | 79 | # Step 3: Check document processing status 80 | print("\nStep 3: Checking document processing status") 81 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/documents/{batch}/indexing-status" 82 | 83 | print(f"Request URL: {url}") 84 | 85 | response = requests.get(url, headers=headers) 86 | print(f"Response status code: {response.status_code}") 87 | print(f"Response content: {response.text}") 88 | 89 | if response.status_code == 200: 90 | result = response.json() 91 | documents = result.get('data', []) 92 | if documents: 93 | document = documents[0] 94 | status = document.get('indexing_status', 'unknown') 95 | print(f"Document status: {status}") 96 | 97 | # 测试输出JSON响应 98 | print("\nStep 4: 输出最终JSON响应") 99 | final_response = { 100 | "status": "success", 101 | "knowledge_base_id": dataset_id, # 单独列出知识库ID作为顶级字段 102 | "knowledge_base": { 103 | "id": dataset_id, 104 | "name": knowledge_base_name 105 | }, 106 | "document": { 107 | "id": document_id, 108 | "name": document_name, 109 | "batch": batch, 110 | "status": status 111 | } 112 | } 113 | print(f"最终JSON响应: {json.dumps(final_response, indent=2)}") 114 | print(f"知识库ID: {final_response['knowledge_base_id']}") 115 | else: 116 | print("No document status information found") 117 | else: 118 | error_data = response.json() 119 | error_code = error_data.get('code', 'unknown_error') 120 | error_message = error_data.get('message', 'Unknown error') 121 | 122 | if error_code == "archived_document_immutable": 123 | print("Error: The archived document is not editable.") 124 | elif error_code == "document_already_finished": 125 | print("Error: The document has been processed. Please refresh the page or go to the document details.") 126 | elif error_code == "document_indexing": 127 | print("Error: The document is being processed and cannot be edited.") 128 | else: 129 | print(f"Error checking document status: {error_message}") 130 | else: 131 | error_data = response.json() 132 | error_code = error_data.get('code', 'unknown_error') 133 | error_message = error_data.get('message', 'Unknown error') 134 | 135 | if error_code == "no_file_uploaded": 136 | print("Error: Please upload your file.") 137 | elif error_code == "too_many_files": 138 | print("Error: Only one file is allowed.") 139 | elif error_code == "file_too_large": 140 | print("Error: File size exceeded.") 141 | elif error_code == "unsupported_file_type": 142 | print("Error: File type not allowed.") 143 | elif error_code == "high_quality_dataset_only": 144 | print("Error: Current operation only supports 'high-quality' datasets.") 145 | elif error_code == "dataset_not_initialized": 146 | print("Error: The dataset is still being initialized or indexing. Please wait a moment.") 147 | elif error_code == "invalid_metadata": 148 | print("Error: The metadata content is incorrect. Please check and verify.") 149 | else: 150 | print(f"Error creating document: {error_message}") 151 | else: 152 | error_data = response.json() 153 | error_code = error_data.get('code', 'unknown_error') 154 | error_message = error_data.get('message', 'Unknown error') 155 | 156 | if error_code == "dataset_name_duplicate": 157 | print("Error: The dataset name already exists. Please modify your dataset name.") 158 | elif error_code == "invalid_action": 159 | print("Error: Invalid action.") 160 | else: 161 | print(f"Error creating knowledge base: {error_message}") 162 | except Exception as e: 163 | print(f"Error occurred during test: {str(e)}") 164 | 165 | print("Test completed") 166 | 167 | def test_knowledge_upload_tool(): 168 | """Test the KnowledgeUploadTool with the modified output""" 169 | from tools.knowledge_upload import KnowledgeUploadTool 170 | from dify_plugin.entities.tool import ToolInvokeMessage 171 | from unittest.mock import MagicMock, patch 172 | 173 | print("\n=== Testing KnowledgeUploadTool with modified output ===") 174 | 175 | # Create mock objects for runtime and session 176 | mock_runtime = MagicMock() 177 | mock_session = MagicMock() 178 | 179 | # Create tool instance with mock objects 180 | tool = KnowledgeUploadTool(mock_runtime, mock_session) 181 | 182 | # Set test parameters 183 | tool_parameters = { 184 | "knowledge_base_name": "Test KB", 185 | "description": "Test description", 186 | "document_name": "Test Doc", 187 | "text": "This is a test content for the document.", 188 | "permission": "only_me", 189 | "indexing_technique": "high_quality" 190 | } 191 | 192 | # Mock the API calls 193 | tool._create_knowledge_base = lambda headers, name, desc, perm, tech: "test-dataset-id-123" 194 | tool._create_document_by_text = lambda headers, dataset_id, doc_name, text, tech: {"id": "test-doc-id-456", "batch": "test-batch-789"} 195 | tool._check_document_status = lambda headers, dataset_id, batch: "completed" 196 | 197 | # Mock the create_text_message and create_json_message methods 198 | def mock_create_text_message(content): 199 | mock_msg = MagicMock() 200 | mock_msg.message_type = "text" 201 | mock_msg.content = content 202 | return mock_msg 203 | 204 | def mock_create_json_message(content): 205 | mock_msg = MagicMock() 206 | mock_msg.message_type = "json" 207 | mock_msg.content = json.dumps(content) 208 | return mock_msg 209 | 210 | tool.create_text_message = mock_create_text_message 211 | tool.create_json_message = mock_create_json_message 212 | 213 | # Mock environment variable for API key 214 | with patch.dict('os.environ', {'DIFY_KNOWLEDGE_API_KEY': 'test_api_key'}): 215 | # Invoke the tool 216 | messages = list(tool._invoke(tool_parameters)) 217 | 218 | # Print all messages 219 | for i, msg in enumerate(messages): 220 | print(f"Message {i+1}: {msg.message_type}") 221 | if msg.message_type == "json": 222 | json_content = json.loads(msg.content) 223 | print(f"JSON content: {json.dumps(json_content, indent=2)}") 224 | 225 | # Verify the output fields 226 | assert "id" in json_content, "Missing 'id' field in output" 227 | assert "status" in json_content, "Missing 'status' field in output" 228 | assert json_content["status"] == 200, "Status should be 200" 229 | assert json_content["id"] == "test-dataset-id-123", "ID should match the test dataset ID" 230 | 231 | print("✅ Output validation passed!") 232 | else: 233 | print(f"Content: {msg.content}") 234 | 235 | print("=== Test completed ===\n") 236 | 237 | if __name__ == "__main__": 238 | # Run the new test for the modified output 239 | test_knowledge_upload_tool() 240 | 241 | # Uncomment to run the original API test 242 | # try: 243 | # # Step 1: Create knowledge base 244 | # # ... existing code ... 245 | # except Exception as e: 246 | # print(f"Error occurred during test: {str(e)}") 247 | # 248 | # print("Test completed") -------------------------------------------------------------------------------- /test_tool.py.combined: -------------------------------------------------------------------------------- 1 | #!/usr/bin/env python3 2 | -------------------------------------------------------------------------------- /test_tool.py.retrieve: -------------------------------------------------------------------------------- 1 | import os 2 | import json 3 | import requests 4 | 5 | # 测试参数 6 | dataset_id = "your_dataset_id" # 替换为您的知识库ID 7 | query = "测试查询" 8 | search_method = "semantic_search" 9 | reranking_enable = False 10 | top_k = 3 11 | score_threshold_enabled = False 12 | score_threshold = 0.5 13 | 14 | # 设置API Key 15 | api_key = "your_api_key" # 替换为您的API Key 16 | 17 | # 设置请求头 18 | headers = { 19 | 'Authorization': f'Bearer {api_key}', 20 | 'Content-Type': 'application/json' 21 | } 22 | 23 | print("开始API请求测试...") 24 | 25 | # 构建检索模型参数 26 | retrieval_model = { 27 | "search_method": search_method, 28 | "reranking_enable": reranking_enable, 29 | "reranking_mode": None, 30 | "reranking_model": { 31 | "reranking_provider_name": "", 32 | "reranking_model_name": "" 33 | }, 34 | "weights": None, 35 | "top_k": top_k, 36 | "score_threshold_enabled": score_threshold_enabled, 37 | "score_threshold": score_threshold if score_threshold_enabled else None 38 | } 39 | 40 | # 执行知识库检索 41 | print(f"从知识库 {dataset_id} 检索与 '{query}' 相关的信息...") 42 | 43 | try: 44 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/retrieve" 45 | 46 | payload = { 47 | "query": query, 48 | "retrieval_model": retrieval_model 49 | } 50 | 51 | print(f"请求URL: {url}") 52 | print(f"请求参数: {payload}") 53 | 54 | response = requests.post(url, headers=headers, json=payload) 55 | print(f"响应状态码: {response.status_code}") 56 | print(f"响应内容: {response.text}") 57 | 58 | if response.status_code == 200: 59 | result = response.json() 60 | records = result.get('records', []) 61 | 62 | if not records: 63 | print(f"未找到与 '{query}' 相关的信息。") 64 | else: 65 | print(f"找到 {len(records)} 条相关信息:") 66 | 67 | for i, record in enumerate(records): 68 | segment = record.get('segment', {}) 69 | content = segment.get('content', '') 70 | document = segment.get('document', {}) 71 | document_name = document.get('name', '未知文档') 72 | score = record.get('score', 0) 73 | 74 | print(f"结果 {i+1}:") 75 | print(f"文档: {document_name}") 76 | print(f"相关度: {score}") 77 | print(f"内容: {content}") 78 | print("-------------------") 79 | 80 | # 输出最终JSON响应 81 | print("\n最终JSON响应:") 82 | final_response = { 83 | "status": "success", 84 | "query": query, 85 | "knowledge_base_id": dataset_id, 86 | "results": records 87 | } 88 | print(json.dumps(final_response, indent=2, ensure_ascii=False)) 89 | else: 90 | error_data = response.json() 91 | error_code = error_data.get('code', 'unknown_error') 92 | error_message = error_data.get('message', '未知错误') 93 | 94 | if error_code == "dataset_not_found": 95 | print("知识库不存在或无权访问") 96 | elif error_code == "invalid_api_key": 97 | print("API Key无效") 98 | else: 99 | print(f"检索知识库时出错: {error_message}") 100 | except Exception as e: 101 | print(f"测试过程中发生错误: {str(e)}") 102 | 103 | print("测试完成") -------------------------------------------------------------------------------- /test_tool.py.upload: -------------------------------------------------------------------------------- 1 | import os 2 | import json 3 | import time 4 | import requests 5 | 6 | # 测试选项 7 | test_upload = True # 设置为True测试上传功能 8 | test_retrieve = True # 设置为True测试检索功能 9 | 10 | # 上传功能测试参数 11 | knowledge_base_name = "Test KB" 12 | description = "" 13 | document_name = "Test Doc" 14 | text_content = "This is a test content for the document." 15 | permission = "only_me" 16 | indexing_technique = "high_quality" 17 | 18 | # 检索功能测试参数 19 | dataset_id = "your_dataset_id" # 替换为您的知识库ID，如果测试上传成功，会自动使用上传创建的知识库ID 20 | query = "test" 21 | search_method = "semantic_search" 22 | reranking_enable = False 23 | top_k = 3 24 | score_threshold_enabled = False 25 | score_threshold = 0.5 26 | 27 | # 设置API Key 28 | api_key = "test_key" # 替换为您的实际API Key 29 | 30 | # 设置请求头 31 | headers = { 32 | 'Authorization': f'Bearer {api_key}', 33 | 'Content-Type': 'application/json' 34 | } 35 | 36 | print("开始API请求测试...") 37 | 38 | # 上传功能测试 39 | if test_upload: 40 | print("\n===== 测试上传功能 =====") 41 | # 步骤1: 创建知识库 42 | print("步骤1: 创建知识库") 43 | url = "https://api.dify.ai/v1/datasets" 44 | payload = { 45 | "name": knowledge_base_name, 46 | "description": description, 47 | "permission": permission, 48 | "indexing_technique": indexing_technique, 49 | "provider": "vendor" 50 | } 51 | 52 | print(f"请求URL: {url}") 53 | print(f"请求参数: {payload}") 54 | 55 | try: 56 | response = requests.post(url, headers=headers, json=payload) 57 | print(f"响应状态码: {response.status_code}") 58 | print(f"响应内容: {response.text}") 59 | 60 | if response.status_code == 200: 61 | result = response.json() 62 | dataset_id = result.get('id') 63 | print(f"知识库创建成功，ID: {dataset_id}") 64 | 65 | # 步骤2: 通过文本创建文档 66 | print("\n步骤2: 创建文档") 67 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/document/create-by-text" 68 | 69 | # 准备处理规则 70 | process_rule = { 71 | "mode": "automatic" 72 | } 73 | 74 | payload = { 75 | "name": document_name, 76 | "text": text_content, 77 | "indexing_technique": indexing_technique, 78 | "process_rule": process_rule 79 | } 80 | 81 | print(f"请求URL: {url}") 82 | print(f"请求参数: {payload}") 83 | 84 | response = requests.post(url, headers=headers, json=payload) 85 | print(f"响应状态码: {response.status_code}") 86 | print(f"响应内容: {response.text}") 87 | 88 | if response.status_code == 200: 89 | result = response.json() 90 | document = result.get('document', {}) 91 | document_id = document.get('id') 92 | batch = result.get('batch', '') 93 | print(f"文档创建成功，ID: {document_id}, 批次: {batch}") 94 | 95 | # 步骤3: 检查文档处理状态 96 | print("\n步骤3: 检查文档处理状态") 97 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/documents/{batch}/indexing-status" 98 | 99 | print(f"请求URL: {url}") 100 | 101 | response = requests.get(url, headers=headers) 102 | print(f"响应状态码: {response.status_code}") 103 | print(f"响应内容: {response.text}") 104 | 105 | if response.status_code == 200: 106 | result = response.json() 107 | documents = result.get('data', []) 108 | if documents: 109 | document = documents[0] 110 | # Step 1: Create knowledge base 111 | print("Step 1: Creating knowledge base") 112 | url = "https://api.dify.ai/v1/datasets" 113 | payload = { 114 | "name": knowledge_base_name, 115 | "description": description, 116 | "permission": permission, 117 | "indexing_technique": indexing_technique, 118 | "provider": "vendor" 119 | } 120 | 121 | print(f"Request URL: {url}") 122 | print(f"Request parameters: {payload}") 123 | 124 | try: 125 | response = requests.post(url, headers=headers, json=payload) 126 | print(f"Response status code: {response.status_code}") 127 | print(f"Response content: {response.text}") 128 | 129 | if response.status_code == 200: 130 | result = response.json() 131 | dataset_id = result.get('id') 132 | print(f"Knowledge base created successfully, ID: {dataset_id}") 133 | 134 | # Step 2: Create document by text 135 | print("\nStep 2: Creating document") 136 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/document/create-by-text" 137 | 138 | # Prepare processing rules 139 | process_rule = { 140 | "mode": "automatic" 141 | } 142 | 143 | payload = { 144 | "name": document_name, 145 | "text": text_content, 146 | "indexing_technique": indexing_technique, 147 | "process_rule": process_rule 148 | } 149 | 150 | print(f"Request URL: {url}") 151 | print(f"Request parameters: {payload}") 152 | 153 | response = requests.post(url, headers=headers, json=payload) 154 | print(f"Response status code: {response.status_code}") 155 | print(f"Response content: {response.text}") 156 | 157 | if response.status_code == 200: 158 | result = response.json() 159 | document = result.get('document', {}) 160 | document_id = document.get('id') 161 | batch = result.get('batch', '') 162 | print(f"Document created successfully, ID: {document_id}, Batch: {batch}") 163 | 164 | # Step 3: Check document processing status 165 | print("\nStep 3: Checking document processing status") 166 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/documents/{batch}/indexing-status" 167 | 168 | print(f"Request URL: {url}") 169 | 170 | response = requests.get(url, headers=headers) 171 | print(f"Response status code: {response.status_code}") 172 | print(f"Response content: {response.text}") 173 | 174 | if response.status_code == 200: 175 | result = response.json() 176 | documents = result.get('data', []) 177 | if documents: 178 | document = documents[0] 179 | status = document.get('indexing_status', 'unknown') 180 | print(f"Document status: {status}") 181 | 182 | # 测试输出JSON响应 183 | print("\nStep 4: 输出最终JSON响应") 184 | final_response = { 185 | "status": "success", 186 | "knowledge_base_id": dataset_id, # 单独列出知识库ID作为顶级字段 187 | "knowledge_base": { 188 | "id": dataset_id, 189 | "name": knowledge_base_name 190 | }, 191 | "document": { 192 | "id": document_id, 193 | "name": document_name, 194 | "batch": batch, 195 | "status": status 196 | } 197 | } 198 | print(f"最终JSON响应: {json.dumps(final_response, indent=2)}") 199 | print(f"知识库ID: {final_response['knowledge_base_id']}") 200 | else: 201 | print("No document status information found") 202 | else: 203 | error_data = response.json() 204 | error_code = error_data.get('code', 'unknown_error') 205 | error_message = error_data.get('message', 'Unknown error') 206 | 207 | if error_code == "archived_document_immutable": 208 | print("Error: The archived document is not editable.") 209 | elif error_code == "document_already_finished": 210 | print("Error: The document has been processed. Please refresh the page or go to the document details.") 211 | elif error_code == "document_indexing": 212 | print("Error: The document is being processed and cannot be edited.") 213 | else: 214 | print(f"Error checking document status: {error_message}") 215 | else: 216 | error_data = response.json() 217 | error_code = error_data.get('code', 'unknown_error') 218 | error_message = error_data.get('message', 'Unknown error') 219 | 220 | if error_code == "no_file_uploaded": 221 | print("Error: Please upload your file.") 222 | elif error_code == "too_many_files": 223 | print("Error: Only one file is allowed.") 224 | elif error_code == "file_too_large": 225 | print("Error: File size exceeded.") 226 | elif error_code == "unsupported_file_type": 227 | print("Error: File type not allowed.") 228 | elif error_code == "high_quality_dataset_only": 229 | print("Error: Current operation only supports 'high-quality' datasets.") 230 | elif error_code == "dataset_not_initialized": 231 | print("Error: The dataset is still being initialized or indexing. Please wait a moment.") 232 | elif error_code == "invalid_metadata": 233 | print("Error: The metadata content is incorrect. Please check and verify.") 234 | else: 235 | print(f"Error creating document: {error_message}") 236 | else: 237 | error_data = response.json() 238 | error_code = error_data.get('code', 'unknown_error') 239 | error_message = error_data.get('message', 'Unknown error') 240 | 241 | if error_code == "dataset_name_duplicate": 242 | print("Error: The dataset name already exists. Please modify your dataset name.") 243 | elif error_code == "invalid_action": 244 | print("Error: Invalid action.") 245 | else: 246 | print(f"Error creating knowledge base: {error_message}") 247 | except Exception as e: 248 | print(f"Error occurred during test: {str(e)}") 249 | 250 | print("Test completed") -------------------------------------------------------------------------------- /tools/knowledge_retrieve.py: -------------------------------------------------------------------------------- 1 | import os 2 | import json 3 | import requests 4 | from collections.abc import Generator 5 | from typing import Any, Dict, Optional, List 6 | 7 | from dify_plugin import Tool 8 | from dify_plugin.entities.tool import ToolInvokeMessage 9 | 10 | class KnowledgeRetrieveTool(Tool): 11 | def _invoke(self, tool_parameters: dict[str, Any]) -> Generator[ToolInvokeMessage, None, None]: 12 | # Get parameters 13 | dataset_id = tool_parameters.get('dataset_id') 14 | query = tool_parameters.get('query') 15 | search_method = tool_parameters.get('search_method', 'semantic_search') 16 | reranking_enable = tool_parameters.get('reranking_enable', False) 17 | top_k = tool_parameters.get('top_k', 3) 18 | score_threshold_enabled = tool_parameters.get('score_threshold_enabled', False) 19 | score_threshold = tool_parameters.get('score_threshold', 0.5) 20 | 21 | # Debug information 22 | print(f"Received parameters: {tool_parameters}") 23 | 24 | # Check required parameters 25 | if not dataset_id: 26 | yield self.create_text_message("Knowledge base ID is required.") 27 | return 28 | 29 | if not query: 30 | yield self.create_text_message("Query content is required.") 31 | return 32 | 33 | # Get API Key from environment variables 34 | api_key = os.environ.get('DIFY_KNOWLEDGE_API_KEY') 35 | if not api_key: 36 | yield self.create_text_message("API Key not found. Please make sure it's set in the plugin configuration.") 37 | return 38 | 39 | # Set request headers 40 | headers = { 41 | 'Authorization': f'Bearer {api_key}', 42 | 'Content-Type': 'application/json' 43 | } 44 | 45 | # Build retrieval model parameters 46 | retrieval_model = { 47 | "search_method": search_method, 48 | "reranking_enable": reranking_enable, 49 | "reranking_mode": None, 50 | "reranking_model": { 51 | "reranking_provider_name": "", 52 | "reranking_model_name": "" 53 | }, 54 | "weights": None, 55 | "top_k": top_k, 56 | "score_threshold_enabled": score_threshold_enabled, 57 | "score_threshold": score_threshold if score_threshold_enabled else None 58 | } 59 | 60 | # Perform knowledge base retrieval 61 | yield self.create_text_message(f"Retrieving information from knowledge base {dataset_id} related to '{query}'...") 62 | 63 | result = self._retrieve_from_knowledge_base(headers, dataset_id, query, retrieval_model) 64 | if not result: 65 | yield self.create_text_message("Retrieval failed. Please check your API Key and parameters.") 66 | return 67 | 68 | if isinstance(result, str): 69 | # This is an error message 70 | yield self.create_text_message(f"Error: {result}") 71 | return 72 | 73 | records = result.get('records', []) 74 | 75 | if not records: 76 | yield self.create_text_message(f"No information found related to '{query}'.") 77 | return 78 | 79 | # Return retrieval results 80 | yield self.create_text_message(f"Found {len(records)} related results:") 81 | 82 | for i, record in enumerate(records): 83 | segment = record.get('segment', {}) 84 | content = segment.get('content', '') 85 | document = segment.get('document', {}) 86 | document_name = document.get('name', 'Unknown document') 87 | score = record.get('score', 0) 88 | 89 | result_text = f"Result {i+1}:\n" 90 | result_text += f"Document: {document_name}\n" 91 | result_text += f"Relevance: {score}\n" 92 | result_text += f"Content: {content}\n" 93 | result_text += "-------------------" 94 | 95 | yield self.create_text_message(result_text) 96 | 97 | # Return detailed information 98 | yield self.create_json_message({ 99 | "status": "success", 100 | "query": query, 101 | "knowledge_base_id": dataset_id, 102 | "results": records 103 | }) 104 | 105 | def _retrieve_from_knowledge_base(self, headers: Dict, dataset_id: str, query: str, retrieval_model: Dict) -> Optional[Dict]: 106 | """Retrieve information from knowledge base""" 107 | try: 108 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/retrieve" 109 | 110 | payload = { 111 | "query": query, 112 | "retrieval_model": retrieval_model 113 | } 114 | 115 | print(f"Knowledge base retrieval request URL: {url}") 116 | print(f"Knowledge base retrieval request parameters: {payload}") 117 | 118 | response = requests.post(url, headers=headers, json=payload) 119 | 120 | print(f"Knowledge base retrieval response status code: {response.status_code}") 121 | print(f"Knowledge base retrieval response content: {response.text}") 122 | 123 | if response.status_code == 200: 124 | return response.json() 125 | else: 126 | error_data = response.json() 127 | error_code = error_data.get('code', 'unknown_error') 128 | error_message = error_data.get('message', 'Unknown error') 129 | 130 | if error_code == "dataset_not_found": 131 | return "Knowledge base does not exist or you don't have access" 132 | elif error_code == "invalid_api_key": 133 | return "Invalid API Key" 134 | else: 135 | print(f"Error retrieving from knowledge base: {error_message}") 136 | print(f"Status code: {response.status_code}") 137 | print(f"Response content: {response.text}") 138 | return f"Retrieval failed: {error_message}" 139 | except Exception as e: 140 | print(f"Error occurred while retrieving from knowledge base: {str(e)}") 141 | return f"Exception occurred: {str(e)}" -------------------------------------------------------------------------------- /tools/knowledge_retrieve.yaml: -------------------------------------------------------------------------------- 1 | identity: 2 | name: knowledge_retrieve 3 | author: stvlynn 4 | label: 5 | en_US: Retrieve from Knowledge Base 6 | zh_Hans: 从知识库检索 7 | description: 8 | human: 9 | en_US: A tool to retrieve information from Dify Knowledge Base. 10 | zh_Hans: 一个从Dify知识库检索信息的工具。 11 | llm: A tool to retrieve information from Dify Knowledge Base. 12 | parameters: 13 | - name: dataset_id 14 | type: string 15 | required: true 16 | label: 17 | en_US: Knowledge Base ID 18 | zh_Hans: 知识库ID 19 | human_description: 20 | en_US: The ID of the knowledge base to retrieve from 21 | zh_Hans: 要检索的知识库ID 22 | llm_description: The ID of the knowledge base to retrieve from 23 | form: form 24 | - name: query 25 | type: string 26 | required: true 27 | label: 28 | en_US: Query 29 | zh_Hans: 查询内容 30 | human_description: 31 | en_US: The query to search for in the knowledge base 32 | zh_Hans: 在知识库中搜索的查询内容 33 | llm_description: The query to search for in the knowledge base 34 | form: llm 35 | - name: search_method 36 | type: select 37 | required: false 38 | options: 39 | - value: keyword_search 40 | label: 41 | en_US: Keyword Search 42 | zh_Hans: 关键词检索 43 | - value: semantic_search 44 | label: 45 | en_US: Semantic Search 46 | zh_Hans: 语义检索 47 | - value: full_text_search 48 | label: 49 | en_US: Full Text Search 50 | zh_Hans: 全文检索 51 | - value: hybrid_search 52 | label: 53 | en_US: Hybrid Search 54 | zh_Hans: 混合检索 55 | default: semantic_search 56 | label: 57 | en_US: Search Method 58 | zh_Hans: 搜索方法 59 | human_description: 60 | en_US: The method to use for searching the knowledge base 61 | zh_Hans: 用于搜索知识库的方法 62 | llm_description: The method to use for searching the knowledge base 63 | form: form 64 | - name: reranking_enable 65 | type: boolean 66 | required: false 67 | default: false 68 | label: 69 | en_US: Enable Reranking 70 | zh_Hans: 启用重排序 71 | human_description: 72 | en_US: Whether to enable reranking of search results 73 | zh_Hans: 是否启用搜索结果重排序 74 | llm_description: Whether to enable reranking of search results 75 | form: form 76 | - name: top_k 77 | type: number 78 | required: false 79 | default: 3 80 | label: 81 | en_US: Number of Results 82 | zh_Hans: 结果数量 83 | human_description: 84 | en_US: The number of results to return 85 | zh_Hans: 返回的结果数量 86 | llm_description: The number of results to return 87 | form: form 88 | - name: score_threshold_enabled 89 | type: boolean 90 | required: false 91 | default: false 92 | label: 93 | en_US: Enable Score Threshold 94 | zh_Hans: 启用分数阈值 95 | human_description: 96 | en_US: Whether to enable score threshold filtering 97 | zh_Hans: 是否启用分数阈值过滤 98 | llm_description: Whether to enable score threshold filtering 99 | form: form 100 | - name: score_threshold 101 | type: number 102 | required: false 103 | default: 0.5 104 | label: 105 | en_US: Score Threshold 106 | zh_Hans: 分数阈值 107 | human_description: 108 | en_US: The minimum score threshold for results (0-1) 109 | zh_Hans: 结果的最小分数阈值（0-1） 110 | llm_description: The minimum score threshold for results (0-1) 111 | form: form 112 | extra: 113 | python: 114 | source: tools/knowledge_retrieve.py -------------------------------------------------------------------------------- /tools/knowledge_upload.py: -------------------------------------------------------------------------------- 1 | import os 2 | import json 3 | import time 4 | import requests 5 | from collections.abc import Generator 6 | from typing import Any, Dict, Optional, List 7 | 8 | from dify_plugin import Tool 9 | from dify_plugin.entities.tool import ToolInvokeMessage 10 | 11 | class KnowledgeUploadTool(Tool): 12 | def _invoke(self, tool_parameters: dict[str, Any]) -> Generator[ToolInvokeMessage, None, None]: 13 | # Get parameters 14 | knowledge_base_name = tool_parameters.get('knowledge_base_name') 15 | description = tool_parameters.get('description', '') 16 | document_name = tool_parameters.get('document_name') 17 | text_content = tool_parameters.get('text') 18 | permission = tool_parameters.get('permission', 'only_me') 19 | indexing_technique = tool_parameters.get('indexing_technique', 'high_quality') 20 | 21 | # Debug information 22 | print(f"Received parameters: {tool_parameters}") 23 | print(f"Text content: {text_content}") 24 | 25 | # Check required parameters 26 | if not knowledge_base_name: 27 | yield self.create_text_message("Knowledge base name is required.") 28 | return 29 | 30 | if not document_name: 31 | yield self.create_text_message("Document name is required.") 32 | return 33 | 34 | if not text_content: 35 | yield self.create_text_message("Text content is required.") 36 | return 37 | 38 | # Get API Key from environment variables 39 | api_key = os.environ.get('DIFY_KNOWLEDGE_API_KEY') 40 | if not api_key: 41 | yield self.create_text_message("API Key not found. Please make sure it's set in the plugin configuration.") 42 | return 43 | 44 | # Set request headers 45 | headers = { 46 | 'Authorization': f'Bearer {api_key}', 47 | 'Content-Type': 'application/json' 48 | } 49 | 50 | # Step 1: Create knowledge base 51 | yield self.create_text_message(f"Creating knowledge base: {knowledge_base_name}...") 52 | 53 | dataset_id = self._create_knowledge_base(headers, knowledge_base_name, description, permission, indexing_technique) 54 | if not dataset_id: 55 | yield self.create_text_message("Failed to create knowledge base. Please check your API Key and parameters.") 56 | return 57 | 58 | yield self.create_text_message(f"Knowledge base created successfully, ID: {dataset_id}") 59 | 60 | # Step 2: Create document with text 61 | yield self.create_text_message(f"Creating document: {document_name}...") 62 | 63 | document_result = self._create_document_by_text(headers, dataset_id, document_name, text_content, indexing_technique) 64 | if not document_result: 65 | yield self.create_text_message("Failed to create document. Please check your parameters.") 66 | return 67 | 68 | if isinstance(document_result, str): 69 | # This is an error message 70 | yield self.create_text_message(f"Error: {document_result}") 71 | return 72 | 73 | document_id = document_result.get('id') 74 | batch = document_result.get('batch') 75 | 76 | yield self.create_text_message(f"Document created successfully, ID: {document_id}, Batch: {batch}") 77 | 78 | # Step 3: Check document processing status 79 | yield self.create_text_message(f"Processing document, please wait...") 80 | 81 | status_result = self._check_document_status(headers, dataset_id, batch) 82 | 83 | if isinstance(status_result, str) and status_result.startswith("Error:"): 84 | yield self.create_text_message(status_result) 85 | return 86 | 87 | status = status_result 88 | 89 | if status == "completed": 90 | yield self.create_text_message(f"Document processing completed!") 91 | elif status == "error" or status == "failed": 92 | yield self.create_text_message(f"Document processing failed. Please check your text content.") 93 | else: 94 | yield self.create_text_message(f"Document is being processed, current status: {status}") 95 | yield self.create_text_message(f"You can check the result later on the Dify platform.") 96 | 97 | # Return detailed information 98 | yield self.create_json_message({ 99 | "status": 200, 100 | "id": dataset_id, 101 | "knowledge_base": { 102 | "id": dataset_id, 103 | "name": knowledge_base_name 104 | }, 105 | "document": { 106 | "id": document_id, 107 | "name": document_name, 108 | "batch": batch, 109 | "status": status 110 | } 111 | }) 112 | 113 | def _create_knowledge_base(self, headers: Dict, name: str, description: str, permission: str, indexing_technique: str) -> Optional[str]: 114 | """Create an empty knowledge base""" 115 | try: 116 | url = "https://api.dify.ai/v1/datasets" 117 | payload = { 118 | "name": name, 119 | "description": description, 120 | "permission": permission, 121 | "indexing_technique": indexing_technique, 122 | "provider": "vendor" 123 | } 124 | 125 | print(f"Knowledge base creation request URL: {url}") 126 | print(f"Knowledge base creation request parameters: {payload}") 127 | 128 | response = requests.post(url, headers=headers, json=payload) 129 | 130 | print(f"Knowledge base creation response status code: {response.status_code}") 131 | print(f"Knowledge base creation response content: {response.text}") 132 | 133 | if response.status_code == 200: 134 | result = response.json() 135 | return result.get('id') 136 | else: 137 | error_data = response.json() 138 | error_code = error_data.get('code', 'unknown_error') 139 | error_message = error_data.get('message', 'Unknown error') 140 | 141 | if error_code == "dataset_name_duplicate": 142 | print(f"Error: The dataset name already exists. Please modify your dataset name.") 143 | elif error_code == "invalid_action": 144 | print(f"Error: Invalid action.") 145 | else: 146 | print(f"Error creating knowledge base: {error_message}") 147 | 148 | print(f"Status code: {response.status_code}") 149 | print(f"Response content: {response.text}") 150 | return None 151 | except Exception as e: 152 | print(f"Error occurred while creating knowledge base: {str(e)}") 153 | return None 154 | 155 | def _create_document_by_text(self, headers: Dict, dataset_id: str, document_name: str, text_content: str, indexing_technique: str) -> Optional[Dict]: 156 | """Create document by text""" 157 | try: 158 | # Correct API endpoint 159 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/document/create-by-text" 160 | 161 | # Prepare processing rules 162 | process_rule = { 163 | "mode": "automatic" 164 | } 165 | 166 | # Ensure text content is a string 167 | if not isinstance(text_content, str): 168 | text_content = str(text_content) 169 | 170 | payload = { 171 | "name": document_name, 172 | "text": text_content, 173 | "indexing_technique": indexing_technique, 174 | "process_rule": process_rule 175 | } 176 | 177 | print(f"Document creation request URL: {url}") 178 | print(f"Document creation request parameters: {payload}") 179 | 180 | response = requests.post(url, headers=headers, json=payload) 181 | 182 | print(f"Document creation response status code: {response.status_code}") 183 | print(f"Document creation response content: {response.text}") 184 | 185 | if response.status_code == 200: 186 | result = response.json() 187 | document = result.get('document', {}) 188 | batch = result.get('batch', '') 189 | return { 190 | 'id': document.get('id'), 191 | 'batch': batch 192 | } 193 | else: 194 | error_data = response.json() 195 | error_code = error_data.get('code', 'unknown_error') 196 | error_message = error_data.get('message', 'Unknown error') 197 | 198 | if error_code == "no_file_uploaded": 199 | return "Error: Please upload your file." 200 | elif error_code == "too_many_files": 201 | return "Error: Only one file is allowed." 202 | elif error_code == "file_too_large": 203 | return "Error: File size exceeded." 204 | elif error_code == "unsupported_file_type": 205 | return "Error: File type not allowed." 206 | elif error_code == "high_quality_dataset_only": 207 | return "Error: Current operation only supports 'high-quality' datasets." 208 | elif error_code == "dataset_not_initialized": 209 | return "Error: The dataset is still being initialized or indexing. Please wait a moment." 210 | elif error_code == "invalid_metadata": 211 | return "Error: The metadata content is incorrect. Please check and verify." 212 | else: 213 | print(f"Error creating document: {error_message}") 214 | print(f"Status code: {response.status_code}") 215 | print(f"Response content: {response.text}") 216 | return None 217 | except Exception as e: 218 | print(f"Error occurred while creating document: {str(e)}") 219 | print(f"Exception details: {repr(e)}") 220 | return None 221 | 222 | def _check_document_status(self, headers: Dict, dataset_id: str, batch: str) -> str: 223 | """Check document processing status""" 224 | try: 225 | # Use the correct API endpoint 226 | url = f"https://api.dify.ai/v1/datasets/{dataset_id}/documents/{batch}/indexing-status" 227 | 228 | print(f"Document status check request URL: {url}") 229 | 230 | # Try to check status, maximum 5 attempts 231 | for i in range(5): 232 | print(f"Document status check attempt {i+1}") 233 | response = requests.get(url, headers=headers) 234 | 235 | print(f"Document status check response status code: {response.status_code}") 236 | print(f"Document status check response content: {response.text}") 237 | 238 | if response.status_code == 200: 239 | result = response.json() 240 | documents = result.get('data', []) 241 | 242 | if documents: 243 | document = documents[0] 244 | status = document.get('indexing_status', 'unknown') 245 | print(f"Document status: {status}") 246 | 247 | if status in ['completed', 'error', 'failed']: 248 | return status 249 | 250 | # If still processing, wait 3 seconds before checking again 251 | print("Document is still being processed, waiting 3 seconds before checking again") 252 | time.sleep(3) 253 | else: 254 | error_data = response.json() 255 | error_code = error_data.get('code', 'unknown_error') 256 | error_message = error_data.get('message', 'Unknown error') 257 | 258 | if error_code == "archived_document_immutable": 259 | return "Error: The archived document is not editable." 260 | elif error_code == "document_already_finished": 261 | return "Error: The document has been processed. Please refresh the page or go to the document details." 262 | elif error_code == "document_indexing": 263 | return "Error: The document is being processed and cannot be edited." 264 | else: 265 | print(f"Error checking document status: {error_message}") 266 | print(f"Response content: {response.text}") 267 | return "Error: Failed to check document status." 268 | 269 | # If still not completed after 5 attempts, return processing status 270 | return "processing" 271 | except Exception as e: 272 | print(f"Error occurred while checking document status: {str(e)}") 273 | return "Error: An unexpected error occurred while checking document status." -------------------------------------------------------------------------------- /tools/knowledge_upload.yaml: -------------------------------------------------------------------------------- 1 | identity: 2 | name: knowledge_upload 3 | author: stvlynn 4 | label: 5 | en_US: Upload to Knowledge Base 6 | zh_Hans: 上传到知识库 7 | description: 8 | human: 9 | en_US: A tool to upload text content to Dify Knowledge Base. 10 | zh_Hans: 一个上传文本内容到Dify知识库的工具。 11 | llm: A tool to upload text content to Dify Knowledge Base. 12 | parameters: 13 | - name: knowledge_base_name 14 | type: string 15 | required: true 16 | label: 17 | en_US: Knowledge Base Name 18 | zh_Hans: 知识库名称 19 | human_description: 20 | en_US: The name of the knowledge base to create 21 | zh_Hans: 要创建的知识库名称 22 | llm_description: The name of the knowledge base to create 23 | form: form 24 | - name: description 25 | type: string 26 | required: false 27 | label: 28 | en_US: Description 29 | zh_Hans: 描述 30 | human_description: 31 | en_US: The description of the knowledge base 32 | zh_Hans: 知识库的描述 33 | llm_description: The description of the knowledge base 34 | form: form 35 | - name: document_name 36 | type: string 37 | required: true 38 | label: 39 | en_US: Document Name 40 | zh_Hans: 文档名称 41 | human_description: 42 | en_US: The name of the document to create 43 | zh_Hans: 要创建的文档名称 44 | llm_description: The name of the document to create 45 | form: form 46 | - name: text 47 | type: string 48 | required: true 49 | label: 50 | en_US: Text Content 51 | zh_Hans: 文本内容 52 | human_description: 53 | en_US: The text content to upload to the knowledge base 54 | zh_Hans: 要上传到知识库的文本内容 55 | llm_description: The text content to upload to the knowledge base 56 | form: llm 57 | - name: permission 58 | type: select 59 | required: true 60 | options: 61 | - value: only_me 62 | label: 63 | en_US: Only Me 64 | zh_Hans: 仅自己 65 | - value: publicly_readable 66 | label: 67 | en_US: Publicly Readable 68 | zh_Hans: 公开可读 69 | default: only_me 70 | label: 71 | en_US: Permission 72 | zh_Hans: 权限 73 | human_description: 74 | en_US: The permission of the knowledge base (only_me or publicly_readable) 75 | zh_Hans: 知识库的权限（仅自己或公开可读） 76 | llm_description: The permission of the knowledge base (only_me or publicly_readable) 77 | form: form 78 | - name: indexing_technique 79 | type: select 80 | required: true 81 | options: 82 | - value: high_quality 83 | label: 84 | en_US: High Quality 85 | zh_Hans: 高质量 86 | - value: economy 87 | label: 88 | en_US: Economy 89 | zh_Hans: 经济 90 | default: high_quality 91 | label: 92 | en_US: Indexing Technique 93 | zh_Hans: 索引技术 94 | human_description: 95 | en_US: The indexing technique to use (high_quality or economy) 96 | zh_Hans: 要使用的索引技术（高质量或经济） 97 | llm_description: The indexing technique to use (high_quality or economy) 98 | form: form 99 | extra: 100 | python: 101 | source: tools/knowledge_upload.py --------------------------------------------------------------------------------