diff --git a/README.md b/README.md
index 556664dd..e0ae70fb 100644
--- a/README.md
+++ b/README.md
@@ -14,12 +14,12 @@ An open-source, self-hosted implementation of the OpenAI Agents API with multipl
 
 ## What it is
 
-OpenAgentCore runs AI agents on your own infrastructure behind the OpenAI Agents API.
+OpenAgentCore runs AI agents on your own infrastructure behind the [OpenAI Agents API](https://developers.openai.com/api/docs/guides/agents-api/overview).
 
-- **Same API as OpenAI.** Point the [official OpenAI SDK](https://developers.openai.com/api/docs/guides/agents/sdk), or plain HTTP, at your installation. No new client to learn.
-- **Your choice of agent.** Each Session runs a native harness: Codex, Claude Code or MiniMax Code, with the model provider you configure.
-- **Your choice of machine.** Agents work in a managed sandbox (Docker, microsandbox or E2B), or on your own Linux, macOS or Windows machine.
-- **Every part is replaceable.** Sandboxes, harnesses and model providers plug in through defined protocols.
+- **Same API as OpenAI.** Point the [OpenAI Agent API](https://developers.openai.com/api/docs/guides/agents/sdk), or plain HTTP, at your installation. No new client to learn.
+- **Your choice of agent.** Each [Session](docs/api/public-agent-api.md) runs a [native harness](contracts/agents-api/harness-onboarding.md): [Codex](https://github.com/openai/codex), [Claude Code](https://code.claude.com/docs/en/overview) or [MiniMax Code](https://github.com/MiniMax-AI/minimax-code), with the [model provider you configure](contracts/agents-api/model-execution.md).
+- **Your choice of machine.** Agents work in a [managed sandbox](contracts/agents-api/sandbox-deployment.md) ([Docker](https://www.docker.com/), [microsandbox](https://github.com/zerocore-ai/microsandbox) or [E2B](https://e2b.dev/)), or on your own Linux, macOS or Windows machine.
+- **Every part is replaceable.** [Sandboxes](docs/sandbox-provider.md), [harnesses](contracts/agents-api/harness-onboarding.md) and [model providers](contracts/agents-api/model-execution.md) plug in through [defined protocols](AGENTS.md#protocols-at-every-boundary).
 
 ## How it fits together
 
diff --git a/README.zh-CN.md b/README.zh-CN.md
index 28e6fe2d..e0df4c5b 100644
--- a/README.zh-CN.md
+++ b/README.zh-CN.md
@@ -27,12 +27,12 @@ OpenAI Agents API 的开源实现，支持多种原生执行引擎，可部署
 
 ## 这是什么
 
-OpenAgentCore 在你自己的基础设施上运行 AI Agent，对外提供 OpenAI Agents API。
+OpenAgentCore 在你自己的基础设施上运行 AI Agent，对外提供 [OpenAI Agents API](https://developers.openai.com/api/docs/guides/agents-api/overview)。
 
 - **与 OpenAI 相同的 API。** [官方 OpenAI SDK](https://developers.openai.com/api/docs/guides/agents/sdk) 或直接 HTTP 调用，改一下地址即可，无需学习新客户端。
-- **自选 Agent。** 每个 Session 运行一个原生 Harness：Codex、Claude Code 或 MiniMax Code， 使用你配置的模型供应商。
-- **自选机器。** Agent 可以在托管沙箱（Docker、microsandbox 或 E2B）里工作， 也可以在你自己的 Linux、macOS 或 Windows 机器上工作。
-- **每个部件都可替换。** 沙箱、Harness 和模型供应商都通过既定协议接入。
+- **自选 Agent。** 每个 [Session](docs/zh/api/public-agent-api.md) 运行一个[原生 Harness](contracts/agents-api/zh/harness-onboarding.md)：[Codex](https://github.com/openai/codex)、[Claude Code](https://code.claude.com/docs/en/overview) 或 [MiniMax Code](https://github.com/MiniMax-AI/minimax-code)， 使用[你配置的模型供应商](contracts/agents-api/zh/model-execution.md)。
+- **自选机器。** Agent 可以在[托管沙箱](contracts/agents-api/zh/sandbox-deployment.md)（[Docker](https://www.docker.com/)、[microsandbox](https://github.com/zerocore-ai/microsandbox) 或 [E2B](https://e2b.dev/)）里工作， 也可以在你自己的 Linux、macOS 或 Windows 机器上工作。
+- **每个部件都可替换。** [沙箱](docs/zh/sandbox-provider.md)、[Harness](contracts/agents-api/zh/harness-onboarding.md) 和[模型供应商](contracts/agents-api/zh/model-execution.md)都通过[既定协议](AGENTS.md#protocols-at-every-boundary)接入。
 
 ## 界面预览