Jentic

MCPAI代理

提供方 jentic

AI代理 HTTP · 远程 会话登录 社区 / 第三方 Claude DesktopClaudeCursorWindsurf
查看接入方式

Jentic SDK & MCP Plugin [Beta] Jentic MCP empowers AI agent builders to discover and integrate external APIs and workflows rapidly—without writing or maintaining API-specific code. This repository contains the core Jentic SDK and the Jentic MCP Plugin. - [Jentic SDK](#jentic-sdk): A comprehensive Python library for discovering and executing APIs and workflows, particularly for LLM tool use.

一句话选型结论

按资料字段整理

Jentic 属于AI代理类 MCP 服务器,HTTP · 远程,会话登录,可直接接入 Claude Desktop、Claude、Cursor 等,有可直接复制的远程端点配置。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类AI代理 传输方式HTTP · 远程 鉴权会话登录 部署方式远程部署
GitHub 星标 未标注 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

AI代理
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

api integrationai agent

安装与配置

HTTP · 远程 会话登录
连远程 HTTPS 端点,通常为 Streamable HTTP
uvx --from git+https://github.com/jentic/jentic-tools.git@main#subdirectory=mcp mcp

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    确认你的客户端支持远程端点,并拿到服务地址。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
给代理接外部能力

让代理在任务中途调用搜索、数据库等真实工具。

多步任务编排

把一串操作拆成步骤,由代理按顺序执行。

自动化重复流程

把固定流程交给代理跑,减少手工操作。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude Desktop、Claude、Cursor,换客户端不用重新研究配置。

先想清楚的地方

  • 远程端点依赖网络与服务方可用性,注意密钥与调用额度。
  • 使用前要准备好会话登录对应的凭据,并按最小权限配置。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在AI代理场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
Jentic当前 未标注 HTTP · 远程 远程部署 有 未标注
Filesystem 560c4363 13 未标注 本地部署 无 2026-05-17
Server Everything (everything) 11 stdio · 本地进程 本地部署 有 2026-05-17
Time 7365cb6d 2 SSE · 服务端推送 远程部署 有 2026-05-18
Everything Claude Code 未标注 stdio · 本地进程 本地部署 有 2026-05-17
Openmemory 557e1585 未标注 stdio · 本地进程 本地部署 有 2026-05-17

服务器档案

AI代理
字段来自项目仓库与公开索引,中文说明由本站标注
提供方jentic上游标注的维护方 jentic
分类AI代理本站按上游分类展示
连接方式HTTP · 远程连远程 HTTPS 端点,通常为 Streamable HTTP
鉴权方式会话登录远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · Cross-platform本地安装时用到的包标识与运行时
部署类型远程部署上游给出的部署形态
最近核验未标注上游最后核验时间

详细介绍

12.4 KB
来自项目仓库与上游目录的原文,未做删改

Jentic Public APIs

> Join our community! Connect with contributors and users on Discord to discuss ideas, ask questions, and collaborate on the Jentic Public APIs repository.
>

Explore the API Directory

Every API has a machine-readable OpenAPI spec in this repository, and most carry an AI-readiness score right alongside it (a scorecard.json next to the spec; see AI-Readiness Scoring below).

Popular APIs

API Category OpenAPI spec
Stripe Payments 📄 OpenAPI
OpenAI AI/ML 📄 OpenAPI
Anthropic (Messages) AI/ML 📄 OpenAPI
GitHub REST Developer Tools 📄 OpenAPI
Slack Web API Communications 📄 OpenAPI
Gmail Productivity 📄 OpenAPI
Notion Productivity 📄 OpenAPI
Shopify Admin E-Commerce 📄 OpenAPI
Plaid Finance 📄 OpenAPI
Discord Communications 📄 OpenAPI
Linear Developer Tools 📄 OpenAPI

Browse by category

The directory is organized into 22 categories. Below is a small sample from a few. Expand a section, then follow View all for every API in that category on jentic.com/apis.

🛠️ Developer Tools

API OpenAPI spec
GitHub REST 📄 OpenAPI
Linear 📄 OpenAPI
Cloudflare 📄 OpenAPI
DigitalOcean 📄 OpenAPI

➡️ View all Developer Tools APIs →

🤖 AI/ML

API OpenAPI spec
OpenAI 📄 OpenAPI
Anthropic (Messages) 📄 OpenAPI

➡️ View all AI/ML APIs →

💳 Payments & Finance

API OpenAPI spec
Stripe 📄 OpenAPI
Plaid 📄 OpenAPI
Xero Accounting 📄 OpenAPI

➡️ View all Payments · Finance APIs →

💬 Communications

API OpenAPI spec
Slack Web API 📄 OpenAPI
Discord 📄 OpenAPI
Twilio Messaging 📄 OpenAPI
SendGrid Mail 📄 OpenAPI

➡️ View all Communications APIs →

📋 Productivity

API OpenAPI spec
Gmail 📄 OpenAPI
Google Calendar 📄 OpenAPI
Google Sheets 📄 OpenAPI
Notion 📄 OpenAPI

➡️ View all Productivity APIs →

Fun & niche APIs

The directory goes well beyond the usual suspects. Here's a taste of the long tail:

API What it's for OpenAPI spec
FaceCheck.ID Reverse face search 📄 OpenAPI
Beatport Electronic music catalog 📄 OpenAPI
Untappd Beer check-ins & ratings 📄 OpenAPI
PokéAPI Pokémon data 📄 OpenAPI
NASA APOD Astronomy Picture of the Day 📄 OpenAPI
MusicBrainz Open music metadata 📄 OpenAPI
Open Brewery DB Brewery directory 📄 OpenAPI
TVmaze TV show schedules & data 📄 OpenAPI
Chess.com Players, games & stats 📄 OpenAPI
Lichess Open chess platform 📄 OpenAPI

> Looking for something else? Search the full API Directory →, or jump into the raw specs via the Quick access API Index above.

Overview

The agentic knowledge layer

AI agents depend on APIs. Their capabilities are defined by the APIs they know about, and their reliability is defined by the quality of that knowledge. Documentation was previously nice-to-have, but for AI it's a need-to-have. The goal of this project is to collate all knowledge about all the world's APIs into a communal, detailed, comprehensive, structured documentation catalog designed for use by AI. This allows AI to accurately generate API integration code, and it allows agents to plan and interact with APIs reliably, without intermediaries.

Open source and open standards

This communal effort requires a stable but extensible representation format that can describe all salient aspects of APIs and associated workflows in full detail. The OpenAPI specifications provide the de-facto standard for formal API descriptions, are widely adopted, supported by a vast ecosystem of associated tooling, and governed by the Linux Foundation. Importantly, the OpenAPI Initiative's most recent specification, Arazzo, allows complex multi-API workflows to be described in a declarative format.

The Jentic API Directory is an open-source catalog of API and workflow descriptions that builds upon these open standards to contribute API and workflow knowledge to AI agents. We will coordinate with the OpenAPI community, and propose an RFC containing various extensions to capture additional knowledge that is especially relevant in the context of AI agents (for example concerning authentication, rate limiting, pricing, governance and safety). If you have suggestions to improve the directory, we welcome discussion on our Discord and PRs on this repository.

AI-scale

Documenting all the world's API knowledge is made achievable by generative AI. Our starting point was OpenAPI documents provided by various vendors online (with special credit to the APIs.guru repository). On top of this, we have generated thousands Arazzo workflows using AI. We are growing this repository using AI agents to import (and improve) existing OpenAPI documents and to generate new OpenAPI specifications where no structured documentation previously existed. Our AI agents are also discovering novel Arazzo workflows that can be performed on top of that API knowledge. We will propose a scorecard evaluation to measure the quality of the generated documentation, allowing us to ensure that both the quantity and the quality of documentation increases as we progress.

We welcome all contributions from the community and from partners who want to accelerate this effort with their own resources and ingenuity. We will ensure that all contributions help the knowledge about each API and workflow in the repository to converge on the best canonical version.

Repository Focus

The repository focuses on:
1. Standardized OpenAPI specifications for public APIs
2. Arazzo workflows that define composable operations across one or more APIs
3. Associated tooling, for example to help import and enrich documentation, or to convert it out into other formats (e.g., AI model provider's tool definition formats).
4. Evaluations and scorecards to measure API knowledge completeness, accuracy and AI-readiness
5. RFCs for extensions to open formats used in the repository, and any other proposals.

Project Stage

> Note: This project is currently in ALPHA.

Documentation

  • STRUCTURE.md - The standardized repository structure for the Jentic API Directory
  • FEEDBACK-FILES.md - Documentation of feedback.json files that track API specification repairs
  • CONTRIBUTING.md - Guidelines for contributing to the repository
  • CODE_OF_CONDUCT.md - Community standards and expectations
  • LICENSE.md - CC0 1.0 License for this repository

Repository Structure

This repository follows a standardized directory structure for organizing API specifications and workflows.

  • API specifications are organized by vendor and version
  • Workflows are organized to clearly show which APIs they reference
  • Multi-API workflows demonstrate how different services can be orchestrated together

For detailed information, please refer to the structure documentation.

Run any of these APIs with Jentic One

The specs in this directory aren't just for reading. An AI agent can execute any of them through Jentic One, a self-hosted execution layer for AI agents. You register the APIs an agent may use and store the credentials once; the agent makes its calls through the credential-injecting Broker, which checks the agent's permissions, attaches the credential at execution time, and logs the call. Your agent never sees your keys, and nothing leaves your own infrastructure.

Stand up an instance and install the CLI by following the Jentic One Quickstart, then walk through making your first brokered call.

With a running instance, point the jentic CLI at any vendor in this catalog and call it. The same steps are available through the Jentic One UI, or by an agent calling the jentic CLI tools directly:```bash
jentic catalog import stripe.com # import an API from this directory (auto-promotes to live)
jentic search "create a payment" # find the right operation
jentic inspect <operation> # see its method, params, and schemas
jentic execute <operation> --json # call it through the Broker


MCP-capable agents don&#x27;t need the CLI in the loop. Jentic One also speaks **MCP**: a local `jentic mcp` server over stdio, or a Streamable HTTP endpoint at `/mcp`, exposes the same discover → execute loop as tools (`search_apis`, `inspect_operation`, `execute`), with every call still routed through the Broker. See connecting an agent for the MCP and raw-HTTP paths.

Any of the thousands of APIs in this repository can be imported and executed this way.

## AI-Readiness Scoring

OpenAPI documents in this repository can be scored for AI-readiness using the **Jentic API Scorecard CLI**, with no signup, no API key, and no configuration needed.

For most APIs the score is already computed and stored in the repository as a `scorecard.json` file sitting next to the spec, so you can read it without running anything:```
apis/openapi/{vendor}/{api-name}/{version}/
├── openapi.json      # the API spec
└── scorecard.json    # its AI-readiness score

URL format

Files follow a predictable raw URL pattern:```
https://raw.githubusercontent.com/jentic/jentic-public-apis/refs/heads/main/apis/openapi/{vendor}/{api-name}/{version}/openapi.json


### Running a score

Pass any spec URL directly to the CLI:```bash
npx @jentic/api-scorecard-cli@latest score \
  https://raw.githubusercontent.com/jentic/jentic-public-apis/refs/heads/main/apis/openapi/swagger-api/petstore/1.0.27/openapi.json

To learn more about the API Scoring Cli and Jentic API AI-Readiness Framework, visit the Jentic API Scorecard CLI repository.

Acknowledgments

The Jentic Public APIs project is built upon the foundation of open standards and community contributions. We extend our sincere gratitude to:

The OpenAPI Initiative and Linux Foundation - For creating, maintaining, and governing the OpenAPI Specification and Arazzo standards that form the backbone of this project. The OpenAPI Initiative's commitment to open standards, extensive ecosystem of tooling, and collaborative governance model directly enable our vision of a communal API knowledge layer. Special recognition goes to the technical steering committee and contributors who have developed these specifications to serve as the de-facto standard for API documentation, making structured, machine-readable API descriptions possible at scale.

APIs.guru - For providing the initial collection of OpenAPI specifications that helped bootstrap this repository. Their dedication to cataloging public APIs has been instrumental in our mission to create an open knowledge foundation for AI agents, and their pioneering work in API discovery laid important groundwork for projects like ours.

The broader API community - Including all the API providers, documentation authors, and open source contributors whose work makes comprehensive API knowledge possible. This project represents a continuation of the collaborative spirit that has driven API standardization and tooling development.

We are committed to contributing back to these communities through our proposed RFCs, quality improvements to existing specifications, and by demonstrating new applications of these open standards in the context of AI agents.

Contributing

We welcome contributions from the community! Whether you're enhancing existing API specifications, creating new Arazzo workflows, or improving documentation, your contributions help build the open knowledge foundation for AI agents.

Please read our Contributing Guidelines for more information on how to get started.

License

This project is licensed under the CC0 1.0 License - see the LICENSE.md file for details.

> Disclaimer: API specifications and workflows in this repository are based on publicly documented third-party APIs, with some modifications such as repairs to OpenAPI specs or creation of new specs from public API documentation. All trademarks and service marks are the property of their respective owners. This repository does not grant rights to use the underlying APIs.

Star History

常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
Jentic 怎么安装?

在客户端按下面的安装命令接入:uvx --from git+https://github.com/jentic/jentic-tools.git@main#subdirectory=mcp mcp。装完在客户端配置里引用同一个包或端点。

Jentic 怎么在 Claude Desktop、Claude、Cursor 里配置?

它的传输方式是HTTP · 远程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

Jentic 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

Jentic 需要 API Key 吗?

资料标注需要会话登录认证,使用前准备好对应凭据,并按最小权限配置。

Jentic 支持哪些客户端?

资料标注支持 Claude Desktop、Claude、Cursor、Windsurf、Codex、Zed;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

Jentic 能提供哪些能力?

资料显示能力标签包括 api integration、ai agent。具体工具清单以官方文档为准。

Jentic 需要什么运行环境?

运行时为 Cross-platform;属于远程端点,不需要在本地装运行时。

Jentic 是本地还是远程?

它是远程端点,连服务地址即可。传输方式为HTTP · 远程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 未标注。