Akeyless MCP Interop (OpenID AIIM)

Akeyless MCP Interop (OpenID AIIM)

Implements an MCP server with OAuth 2.1 Protected Resource Metadata, enabling token-based authentication for MCP tools like ping.

Category
访问服务器

README

Akeyless MCP Interop (OpenID AIIM)

HTTP MCP Server for the OpenID Foundation AIIM MCP Security Interoperability event (Gartner IAM Summit 2026).

Item Value
Formal role MCP Server only
Target matrix cell OAuth 2.1 — OPRM (OAuth Protected Resource Metadata)
Out of initial scope CIMD, EMA/ID-JAG, MCP Gateway
Booth ARA is demonstrated on the booth (separate from this repo)

Partners verify that this server advertises OPRM and accepts Bearer access tokens. That qualifies the OAuth token-based access × MCP Server cell.


Who continues from here

If you are picking this up:

  1. Read this README.
  2. Run Local quick start below and confirm tests pass.
  3. Follow docs/DEPLOY-AWS.md to put a public HTTPS endpoint online.
  4. Follow docs/TRACK-A-DEMO.md with a partner MCP Client.
  5. Fill the OAuth matrix (each side fills the other role column) and submit to AIIM CG — first success due 16 Oct 2026.
  6. Commit email to OpenID (if not already sent) is due 10 Aug 2026 PT — see docs/HANDOFF.md.

What this server does

Endpoint Auth Purpose
GET /.well-known/oauth-protected-resource none RFC 9728 OPRM document
GET /health none Liveness
POST /mcp Bearer JWT Streamable HTTP MCP (ping tool)

Unauthenticated /mcp calls return 401 with:

WWW-Authenticate: Bearer realm="akeyless-openid-mcp", resource_metadata="<oprm-url>", scope="mcp:tools"

Token validation uses either:

  • INTEROP_JWKS_URI (preferred for partner tests), or
  • INTEROP_JWT_SECRET (HS256 local/demo only — do not use for official partner AS tests)

Local quick start

Requires Node.js 18+.

git clone https://github.com/akeyless-community/mcp-interop.git
cd mcp-interop
npm install
npm test
npm run build

cp .env.example .env
# edit .env if needed, then:
set -a && source .env && set +a
npm start

Smoke checks (another terminal):

curl -sS http://127.0.0.1:8790/health
curl -sS http://127.0.0.1:8790/.well-known/oauth-protected-resource
curl -sS -D - -o /dev/null -X POST http://127.0.0.1:8790/mcp \
  -H 'Content-Type: application/json' -d '{}'

Expect OPRM JSON and a 401 with resource_metadata + scope on the POST.


Partner / production config

Set public HTTPS URLs (example hostname):

export INTEROP_HTTP_HOST=127.0.0.1
export INTEROP_HTTP_PORT=8790
export INTEROP_HTTP_PATH=/mcp
export INTEROP_RESOURCE=https://mcp-interop.example.com/mcp
export INTEROP_PUBLIC_BASE_URL=https://mcp-interop.example.com
export INTEROP_RESOURCE_METADATA_URL=https://mcp-interop.example.com/.well-known/oauth-protected-resource
export INTEROP_SCOPES=mcp:tools
export INTEROP_AUTHORIZATION_SERVERS=https://partner-as.example.com
export INTEROP_JWT_ISSUER=https://partner-as.example.com
export INTEROP_JWT_AUDIENCE=https://mcp-interop.example.com/mcp
export INTEROP_JWKS_URI=https://partner-as.example.com/.well-known/jwks.json
# do not set INTEROP_JWT_SECRET for partner tests

Full AWS / nginx / systemd steps: docs/DEPLOY-AWS.md.


Give partners

Item Value
MCP Server URL https://<your-dns>/mcp
OPRM URL https://<your-dns>/.well-known/oauth-protected-resource
Required scope mcp:tools
Resource / audience https://<your-dns>/mcp

Partner flow: call without token → read resource_metadata → get token from AS listed in OPRM → call /mcp with Authorization: Bearer <token> → ping.


Repo layout

src/
  index.ts      # process entry
  http.ts       # Express + Streamable HTTP + OPRM routes + auth gate
  oprm.ts       # RFC 9728 document + WWW-Authenticate builder
  auth.ts       # Bearer JWT verification (JWKS or HS256)
  config.ts     # INTEROP_* env loading
  server.ts     # MCP Server + ping tool
  oprm.test.ts  # unit tests
docs/
  HANDOFF.md      # status, deadlines, next owner checklist
  DEPLOY-AWS.md   # public HTTPS hosting
  TRACK-A-DEMO.md # partner OAuth matrix test script

Scripts

Command Purpose
npm test Unit tests
npm run build Compile to dist/
npm start Run dist/index.js

References

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选