總覽Overview
三個環境:stg / cl1 / prodThree environments: stg / cl1 / prod
ContactLoop 有三個獨立部署的環境,各自有獨立的 SW 與 Ava、獨立的帳號與 Account ID。新人一律先在 stg 練習。ContactLoop has three independently deployed environments, each with its own SW and Ava, its own accounts and Account IDs. New hires always start in stg.
| 環境Env | SW Manager URL | Ava Messenger URL | 用途 / 風險Use / risk |
|---|---|---|---|
| stg Staging | staging.sw.onramplab.com |
staging.ava.onramplab.com |
開發整合測試、QA 跑測試的主要環境。🟢 低風險,可亂搞。Dev integration testing and where QA runs tests. 🟢 Low risk — break things freely. |
| cl1 CL1 | manager-cl1.contactloop.com |
messenger-cl1.contactloop.com |
Pre-prod / 客戶預覽。🟡 中風險,部分新版頁面可能還沒跟上 stg。Pre-prod / customer preview. 🟡 Medium risk — some new pages may lag behind stg. |
| prod Production | manager.contactloop.com |
messenger.contactloop.com |
正式流量、真實客戶資料。🔴 高風險,勿任意匯入或觸發。Live traffic, real customer data. 🔴 High risk — do not import or trigger carelessly. |
每個環境的 Account ID 都不同。Each environment has a different Account ID.例如 stg 的 sw_nucleus.account_id=93,cl1 是 38,prod 是 361。把 stg 的設定直接搬到 prod 一定對不上——所有設定以 conftest.py 的 ENV_CONFIG 為準。 For example, stg has sw_nucleus.account_id=93, cl1 is 38, prod is 361. Copying stg settings straight to prod will never match — all settings follow ENV_CONFIG in conftest.py.
環境版本會落差(常見困惑來源)Environments can drift (a common source of confusion)
stg 通常是最新版,cl1 / prod 的前端部署會晚一點跟上。例如 2026-08 實測:stg 的 Contacts 系列已改打 /api/contacts/grid,但 cl1 的 Marketing Sources / Endpoint Attributes / Technographics 三頁還在打舊版 /api/contacts?…size=500。測試裡這類情況會標成 xfail(預期失敗),部署跟上後自動轉 XPASS。看到 xfail 不是壞掉,是「已知環境落差」。stg is usually the newest; cl1 / prod frontend deploys follow a bit later. Example (measured 2026-08): stg's Contacts pages already call /api/contacts/grid, but cl1's Marketing Sources / Endpoint Attributes / Technographics still call the old /api/contacts?…size=500. In tests these are marked xfail (expected fail) and flip to XPASS once the deploy catches up. An xfail isn't a breakage — it's a "known environment drift."