官方说明2 个来源昨天
WorkGPT
GPT agent framework for invoking APIs
- 来源明细
- 收录自 e2b-dev/awesome-ai-agents 榜单(我们从中收集的条目之一) · 条目类型:开源项目
- 收录与活跃
- 收录于 2026-10-16 · 最近见到 2026-10-16 · 共见到 3 次 · 被 2 个来源列出
- 评分构成
- 综合分 203 · 2 个来源 · repo
常见问题(答案都来自我们的记录)
怎么安装?
我们【没有】从该仓库取到安装命令(我们的 install 覆盖率只有约 4%)。请看它的 README 或 releases 页。
它适合谁 / 用来做什么?
我们还没有为它记录标签。
谁维护它、还在更新吗?
我们收录于 2026-10-16;最近见到 2026-10-16。这里没有"官方认证"或"质量评级"——我们只记录来源与时间。
workgpt
以下摘录自该仓库的 README(片段) · 在 GitHub 读完整 README →
WorkGPT is an agent framework in a similar fashion to AutoGPT or LangChain. You give it a directive and an array of APIs and it will converse back and forth with the AI until its directive is complete.
For example, a directive could be to research the web for something, to crawl a website, or to order you an Uber. We support any and all APIs that can be represented with an OpenAPI file.
Install
Usage
What is OpenPM
You'll notice that we're using an OpenPM API in the example above. OpenPM is a package manager for OpenAPI files. In the example you can see we've pulled in a package from OpenPM called ipinfo to be used for looking up IP addresses.
You don't have to use OpenPM. We support importing any arbitrary OpenAPI file. You can see that we're smart about authentication. You just need to pass an authKey and the library will figure out how to authorize itself. All the endpoints in the API will be exposed as local functions to the LLM, ready to be invoked.
Crawling example
We include an example of using Puppeteer as a text-based browser to give the LLM access to the web. You'll notice that the text-based browser passes out all the HTML and just returns text. This is actually enough for GPT-4, which is smart enough to extract data from the plain text.
We can pass our own custom API to the LLM as a finishing program API that the LLM can call whenever it's finished. The advantage of this is that you can give it a schema to follow, which is great when trying to extract data from a webpage.
License
MIT
- 来源
- awesome
· https://github.com/team-openpm/workgpt
- 收录于
- 昨天
← 返回索引