Github 工具包
该 Github
工具包包含使大型语言模型代理能够与 GitHub 仓库交互的工具。
该工具是 PyGitHub 库的封装。
有关所有 GithubToolkit 功能和配置的详细文档,请访问 API 参考。
设置
从高层次来看,我们将:
- 安装 pygithub 库
- 创建一个 Github 应用
- 设置您的环境变量
- 使用
toolkit.get_tools()
将工具传递给您的代理
如果您想从单个工具的运行中获取自动化跟踪,您还可以通过取消下面的注释来设置您的LangSmith API 密钥:
# os.environ["LANGSMITH_API_KEY"] = getpass.getpass("Enter your LangSmith API key: ")
# os.environ["LANGSMITH_TRACING"] = "true"
安装
1. 安装依赖
此集成在 langchain-community
中实现。我们还需要 pygithub
依赖:
%pip install --upgrade --quiet pygithub langchain-community
2. 创建一个 Github 应用
按照这里的说明 创建并注册一个 Github 应用。确保您的应用具有以下 仓库权限:
- 提交状态(只读)
- 内容(读写)
- 问题(读写)
- 元数据(只读)
- 拉取请求(读写)
一旦应用程序注册完成,您必须授予应用程序访问您希望其操作的每个存储库的权限。请使用 github.com 这里的应用设置。
3. 设置环境变量
在初始化您的代理之前,需要设置以下环境变量:
- GITHUB_APP_ID - 在您应用程序的一般设置中找到的六位数字
- GITHUB_APP_PRIVATE_KEY - 您应用程序的私钥 .pem 文件的位置,或该文件的完整文本作为字符串。
- GITHUB_REPOSITORY - 您希望机器人操作的Github存储库的名称。必须遵循格式 {username}/{repo-name}。确保该应用程序已首先添加到此存储库!
- 可选: GITHUB_BRANCH - 机器人将进行提交的分支。默认为
repo.default_branch
。 - 可选: GITHUB_BASE_BRANCH - 您的存储库的基础分支,PR 将基于此。默认为
repo.default_branch
。
import getpass
import os
for env_var in [
"GITHUB_APP_ID",
"GITHUB_APP_PRIVATE_KEY",
"GITHUB_REPOSITORY",
]:
if not os.getenv(env_var):
os.environ[env_var] = getpass.getpass()
实例化
现在我们可以实例化我们的工具包:
<!--IMPORTS:[{"imported": "GitHubToolkit", "source": "langchain_community.agent_toolkits.github.toolkit", "docs": "https://python.langchain.com/api_reference/community/agent_toolkits/langchain_community.agent_toolkits.github.toolkit.GitHubToolkit.html", "title": "Github Toolkit"}, {"imported": "GitHubAPIWrapper", "source": "langchain_community.utilities.github", "docs": "https://python.langchain.com/api_reference/community/utilities/langchain_community.utilities.github.GitHubAPIWrapper.html", "title": "Github Toolkit"}]-->
from langchain_community.agent_toolkits.github.toolkit import GitHubToolkit
from langchain_community.utilities.github import GitHubAPIWrapper
github = GitHubAPIWrapper()
toolkit = GitHubToolkit.from_github_api_wrapper(github)
工具
查看可用工具:
tools = toolkit.get_tools()
for tool in tools:
print(tool.name)
Get Issues
Get Issue
Comment on Issue
List open pull requests (PRs)
Get Pull Request
Overview of files included in PR
Create Pull Request
List Pull Requests' Files
Create File
Read File
Update File
Delete File
Overview of existing files in Main branch
Overview of files in current working branch
List branches in this repository
Set active branch
Create a new branch
Get files from a directory
Search issues and pull requests
Search code
Create review request
这些工具的目的如下:
下面将详细解释每个步骤。
-
获取问题- 从仓库中获取问题。
-
获取问题- 获取特定问题的详细信息。
-
对问题评论- 在特定问题上发布评论。
-
创建拉取请求- 从机器人的工作分支创建一个拉取请求到基础分支。
-
创建文件- 在仓库中创建一个新文件。
-
读取文件- 从仓库中读取文件。
-
更新文件 - 更新存储库中的文件。
-
删除文件 - 从存储库中删除文件。
在代理中使用
我们需要一个大型语言模型或聊天模型:
- OpenAI
- Anthropic
- Azure
- Cohere
- NVIDIA
- FireworksAI
- Groq
- MistralAI
- TogetherAI
pip install -qU langchain-openai
import getpass
import os
os.environ["OPENAI_API_KEY"] = getpass.getpass()
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(model="gpt-4o-mini")
pip install -qU langchain-anthropic
import getpass
import os
os.environ["ANTHROPIC_API_KEY"] = getpass.getpass()
from langchain_anthropic import ChatAnthropic
llm = ChatAnthropic(model="claude-3-5-sonnet-20240620")
pip install -qU langchain-openai
import getpass
import os
os.environ["AZURE_OPENAI_API_KEY"] = getpass.getpass()
from langchain_openai import AzureChatOpenAI
llm = AzureChatOpenAI(
azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
azure_deployment=os.environ["AZURE_OPENAI_DEPLOYMENT_NAME"],
openai_api_version=os.environ["AZURE_OPENAI_API_VERSION"],
)
pip install -qU langchain-google-vertexai
import getpass
import os
os.environ["GOOGLE_API_KEY"] = getpass.getpass()
from langchain_google_vertexai import ChatVertexAI
llm = ChatVertexAI(model="gemini-1.5-flash")
pip install -qU langchain-cohere
import getpass
import os
os.environ["COHERE_API_KEY"] = getpass.getpass()
from langchain_cohere import ChatCohere
llm = ChatCohere(model="command-r-plus")
pip install -qU langchain-nvidia-ai-endpoints
import getpass
import os
os.environ["NVIDIA_API_KEY"] = getpass.getpass()
from langchain import ChatNVIDIA
llm = ChatNVIDIA(model="meta/llama3-70b-instruct")
pip install -qU langchain-fireworks
import getpass
import os
os.environ["FIREWORKS_API_KEY"] = getpass.getpass()
from langchain_fireworks import ChatFireworks
llm = ChatFireworks(model="accounts/fireworks/models/llama-v3p1-70b-instruct")
pip install -qU langchain-groq
import getpass
import os
os.environ["GROQ_API_KEY"] = getpass.getpass()
from langchain_groq import ChatGroq
llm = ChatGroq(model="llama3-8b-8192")
pip install -qU langchain-mistralai
import getpass
import os
os.environ["MISTRAL_API_KEY"] = getpass.getpass()
from langchain_mistralai import ChatMistralAI
llm = ChatMistralAI(model="mistral-large-latest")
pip install -qU langchain-openai
import getpass
import os
os.environ["TOGETHER_API_KEY"] = getpass.getpass()
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
base_url="https://api.together.xyz/v1",
api_key=os.environ["TOGETHER_API_KEY"],
model="mistralai/Mixtral-8x7B-Instruct-v0.1",
)
用一组工具初始化代理:
from langgraph.prebuilt import create_react_agent
tools = [tool for tool in toolkit.get_tools() if tool.name == "Get Issue"]
assert len(tools) == 1
tools[0].name = "get_issue"
agent_executor = create_react_agent(llm, tools)
并向其发出查询:
example_query = "What is the title of issue 24888?"
events = agent_executor.stream(
{"messages": [("user", example_query)]},
stream_mode="values",
)
for event in events:
event["messages"][-1].pretty_print()
================================[1m Human Message [0m=================================
What is the title of issue 24888?
==================================[1m Ai Message [0m==================================
Tool Calls:
get_issue (call_iSYJVaM7uchfNHOMJoVPQsOi)
Call ID: call_iSYJVaM7uchfNHOMJoVPQsOi
Args:
issue_number: 24888
=================================[1m Tool Message [0m=================================
Name: get_issue
{"number": 24888, "title": "Standardize KV-Store Docs", "body": "To make our KV-store integrations as easy to use as possible we need to make sure the docs for them are thorough and standardized. There are two parts to this: updating the KV-store docstrings and updating the actual integration docs.\r\n\r\nThis needs to be done for each KV-store integration, ideally with one PR per KV-store.\r\n\r\nRelated to broader issues #21983 and #22005.\r\n\r\n## Docstrings\r\nEach KV-store class docstring should have the sections shown in the [Appendix](#appendix) below. The sections should have input and output code blocks when relevant.\r\n\r\nTo build a preview of the API docs for the package you're working on run (from root of repo):\r\n\r\n\`\`\`shell\r\nmake api_docs_clean; make api_docs_quick_preview API_PKG=openai\r\n\`\`\`\r\n\r\nwhere `API_PKG=` should be the parent directory that houses the edited package (e.g. community, openai, anthropic, huggingface, together, mistralai, groq, fireworks, etc.). This should be quite fast for all the partner packages.\r\n\r\n## Doc pages\r\nEach KV-store [docs page](https://python.langchain.com/docs/integrations/stores/) should follow [this template](https://github.com/langchain-ai/langchain/blob/master/libs/cli/langchain_cli/integration_template/docs/kv_store.ipynb).\r\n\r\nHere is an example: https://python.langchain.com/docs/integrations/stores/in_memory/\r\n\r\nYou can use the `langchain-cli` to quickly get started with a new chat model integration docs page (run from root of repo):\r\n\r\n\`\`\`shell\r\npoetry run pip install -e libs/cli\r\npoetry run langchain-cli integration create-doc --name \"foo-bar\" --name-class FooBar --component-type kv_store --destination-dir ./docs/docs/integrations/stores/\r\n\`\`\`\r\n\r\nwhere `--name` is the integration package name without the \"langchain-\" prefix and `--name-class` is the class name without the \"ByteStore\" suffix. This will create a template doc with some autopopulated fields at docs/docs/integrations/stores/foo_bar.ipynb.\r\n\r\nTo build a preview of the docs you can run (from root):\r\n\r\n\`\`\`shell\r\nmake docs_clean\r\nmake docs_build\r\ncd docs/build/output-new\r\nyarn\r\nyarn start\r\n\`\`\`\r\n\r\n## Appendix\r\nExpected sections for the KV-store class docstring.\r\n\r\n\`\`\`python\r\n \"\"\"__ModuleName__ completion KV-store integration.\r\n\r\n # TODO: Replace with relevant packages, env vars.\r\n Setup:\r\n Install ``__package_name__`` and set environment variable ``__MODULE_NAME___API_KEY``.\r\n\r\n .. code-block:: bash\r\n\r\n pip install -U __package_name__\r\n export __MODULE_NAME___API_KEY=\"your-api-key\"\r\n\r\n # TODO: Populate with relevant params.\r\n Key init args \u2014 client params:\r\n api_key: Optional[str]\r\n __ModuleName__ API key. If not passed in will be read from env var __MODULE_NAME___API_KEY.\r\n\r\n See full list of supported init args and their descriptions in the params section.\r\n\r\n # TODO: Replace with relevant init params.\r\n Instantiate:\r\n .. code-block:: python\r\n\r\n from __module_name__ import __ModuleName__ByteStore\r\n\r\n kv_store = __ModuleName__ByteStore(\r\n # api_key=\"...\",\r\n # other params...\r\n )\r\n\r\n Set keys:\r\n .. code-block:: python\r\n\r\n kv_pairs = [\r\n [\"key1\", \"value1\"],\r\n [\"key2\", \"value2\"],\r\n ]\r\n\r\n kv_store.mset(kv_pairs)\r\n\r\n .. code-block:: python\r\n\r\n Get keys:\r\n .. code-block:: python\r\n\r\n kv_store.mget([\"key1\", \"key2\"])\r\n\r\n .. code-block:: python\r\n\r\n # TODO: Example output.\r\n\r\n Delete keys:\r\n ..code-block:: python\r\n\r\n kv_store.mdelete([\"key1\", \"key2\"])\r\n\r\n ..code-block:: python\r\n \"\"\" # noqa: E501\r\n\`\`\`", "comments": "[]", "opened_by": "jacoblee93"}
==================================[1m Ai Message [0m==================================
The title of issue 24888 is "Standardize KV-Store Docs".
API 参考
有关所有 GithubToolkit
功能和配置的详细文档,请访问 API 参考。