archestra README

项目的 README,来自 github.com/archestra-ai/archestra。合并为一个 markdown 文件:README.md。

Archestra

Secure AI Platform

Built on a strong security and observability foundation: SSO and RBAC, sandboxed code execution, Lethal-Trifecta guardrails, OpenTelemetry traces, and Prometheus metrics — first-class, not bolted on.

License Release Contributors Last commit Commit activity

Quickstart  ·  Docs  ·  Releases  ·  Slack


https://github.com/user-attachments/assets/912b85e2-ef14-4e83-a807-5e2a81457a88


What it does

Point your users — or your agents, or Claude / Codex / Cursor — at one URL. Archestra handles the rest:

Already running dangerous single-tenant agents like Claude Cowork, OpenClaw, or Hermes in your enterprise? Migration Kit →

Quickstart

docker pull archestra/platform:latest

docker run \
  -p 127.0.0.1:9000:9000 -p 127.0.0.1:3000:3000 \
  -e ARCHESTRA_QUICKSTART=true \
  -v /var/run/docker.sock:/var/run/docker.sock \
  -v archestra-postgres-data:/var/lib/postgresql/data \
  -v archestra-app-data:/app/data \
  archestra/platform:latest

Open http://localhost:3000. Full Docker / Helm / Kubernetes instructions live in the quickstart docs.

Release channels

Archestra maintains two active release tracks:

New features and bug fixes land on main first and ship in rolling beta releases. Fixes are then selectively backported to the supported stable branch. For production environments, pin an exact version tag or Helm chart version rather than tracking latest. See our Release guide for full release lifecycle details.

Ready for production

Deeper docs

Thank you for continuously making Archestra better — you're awesome 🫶

Contributors

Linux Foundation        CNCF

archestra (archestra-ai/archestra)

正在搜索数据库…

这是什么?Archestra 是一个企业级 AI 平台,集成了 LLM 网关、MCP 注册中心/网关、A2A 网关、智能体运行时和 MCP 编排器,并配有安全防护和可观测性功能。你可以让用户、智能体,或 Claude、Codex、Cursor 等工具,统一通过同一个 URL 接入。
这是单一应用,还是多个网关打包在一起?它是一个自托管平台,内含多个组件:LLM 网关、MCP 网关、A2A 网关、私有 MCP 注册中心、MCP 编排器和智能体运行时,而不是单一用途的工具。
我该怎么运行 Archestra?README 没有直接给出安装命令,而是指向一份专门的 Quickstart 指南和文档,因此部署步骤都在该链接指南中。
要使用 MCP 编排器和智能体运行时,我需要 Kubernetes 吗?需要。MCP 编排器基于 Kubernetes Operator 构建,支持自助式晋升;智能体运行时则使用原生支持 K8s 的文件系统来执行沙箱代码。
这是为谁设计的?面向已经在运行多个 AI 智能体和工具的企业,这些企业需要集中式身份管理、成本控制和安全能力,而不是分散的单租户部署。
我如何防止智能体通过工具调用泄露数据?它针对 'Lethal Trifecta' 模式执行确定性防护规则;MCP 工具调用通过 OAuth On-Behalf-Of 流程转发,使工具以实际用户身份运行,而不是共用的服务账号;所有操作都通过 OpenTelemetry 追踪和 Prometheus 指标记录。
'Lethal Trifecta' 防护旨在防范什么?这是 Archestra 对一种危险组合的称呼:智能体同时具备访问私有数据、接触不可信内容以及对外通信渠道的能力。Archestra 通过确定性防护规则,确保这种组合不会被放任通过。
Cursor 或 Codex 这类编程智能体可以使用 Archestra 吗?可以。Archestra 提供开发者 LLM 与 MCP 门户,通过单一 token/代理,让 Claude Code、Codex 和 Cursor 的调用都经由它转发。
它能同时对接多个 LLM 服务商吗?可以。LLM 网关支持 Anthropic、OpenAI、Azure、Bedrock、DeepSeek 等,并提供成本上限、虚拟 API 密钥和动态模型路由。
用户如何登录 Archestra?通过 SSO(OIDC、SAML、Okta、Entra)登录,并结合 RBAC 角色映射和团队同步进行访问控制。
大小
910.6 MB
星标
4.4K
时长
1年
提交 · 90 天