convex-agent
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinese<!-- GENERATED from convex-agents content/capabilities/agent.json — do not edit by hand. -->
<!-- 由convex-agents content/capabilities/agent.json生成 — 请勿手动编辑。 -->
Add an AI agent / RAG backend
添加AI Agent/RAG后端
Install @convex-dev/agent for durable threads, message history, tool-calls, and vector search/RAG — the backend for an in-app AI agent.
安装@convex-dev/agent以获得持久线程、消息历史记录、工具调用以及向量搜索/RAG功能——这是应用内AI Agent的后端解决方案。
Workflow
工作流程
- Install @convex-dev/agent + add to convex.config.ts.
- Define the agent (model, tools, instructions); store the LLM key via the micro power.
env - Create threads + stream messages; persist history in Convex.
- For RAG: embed docs into a vector index and retrieve in the tool.
- 安装@convex-dev/agent并添加至convex.config.ts。
- 定义Agent（模型、工具、指令）；通过微功能存储LLM密钥。
env - 创建线程并流式传输消息；在Convex中持久化历史记录。
- 针对RAG：将文档嵌入向量索引并在工具中检索。
Rules
规则
- Keep the LLM API key in Convex env (use the micro power), never client-side.
env - Run model calls in actions ('use node' if the SDK needs it).
- Persist threads/messages in Convex for durability + reactivity.
- 将LLM API密钥保存在Convex环境变量中（使用微功能），绝不要放在客户端。
env - 在actions中运行模型调用（如果SDK需要，使用'use node'）。
- 在Convex中持久化线程/消息以实现耐用性和响应性。