````markdown

会话限流快速部署清单¶

(站点快速指南，基于 tools/rag-worker/docs/RATE_LIMITER_QUICKSTART.md)

步骤 1：创建 KV Namespace¶

bash wrangler kv:namespace create "RATE_LIMIT_KV"

步骤 2：更新 wrangler.toml¶

toml [[kv_namespaces]] binding = "RATE_LIMIT_KV" id = "YOUR_ACTUAL_KV_NAMESPACE_ID" preview_id = "YOUR_PREVIEW_KV_ID"

实现注意：限流中间件代码位于 tools/rag-worker/src/middleware/rate-limiter.ts，部署前请确认 tools/rag-worker/src/worker.ts 已导入该中间件并且 wrangler.toml 中 RATE_LIMIT_KV 绑定正确。

步骤 3：配置限流参数（可选）¶

toml [vars] RATE_LIMIT_ENABLED = "true" RATE_LIMIT_WINDOW_MS = "3600000" RATE_LIMIT_MAX_REQUESTS = "20" RATE_LIMIT_COOLDOWN_MS = "900000"

步骤 4：部署并验证¶

```bash npm run deploy

运行测试脚本（示例见 rate-limiter-deployment.md）¶

./test-rate-limit.sh ```

本地单元测试¶

bash npm test tests/unit/rate-limiter.test.ts