docs: distill codex live sessions session

Author: linhay

.agents/skills/gettokens-domain-engineering/SKILL.md

@@ -149,6 +149,7 @@ This skill unifies the technical rules for building, styling, and debugging GetT
   - The frontend must support browser preview with mock snapshots, while desktop mode polls the real Wails snapshot. Show source state such as `live`, `cache`, or `preview` so stale sidecar state is not mistaken for no sessions.
   - Request timing trend charts must be driven by request records, not decorative UI state. Put trend derivation in a pure model under `features/codex-live-sessions/model/`, merge the active request by `requestID`, sort points by `startedAt`, and accept an explicit `nowMs` option so live requests can be projected in tests.
   - For streaming / active / reconnecting requests without `completedAt`, project `totalDurationMs` from `nowMs - startedAt` with a bounded safety cap. Do not mutate the original request or invent first-event / first-token timings.
+  - Only the current active request is allowed to use `nowMs - startedAt` projection in the timing trend. Historical request rows that still carry `streaming` / `reconnecting` from cache or stale sidecar state must keep their recorded `timing.totalDurationMs` or `completedAt - startedAt`; otherwise every total-duration point will grow together.
   - Trend chart x positions should use request timestamps (`startedAtMs` within min/max), not array index spacing. This keeps sparse request bursts and live samples faithful to real time.
   - Keep the live-session chart visually inside the page section, not as a nested card. Use the existing Swiss-industrial chart tokens, footer summaries below the graph, and live markers such as dashed strokes/rings for in-flight samples.
   - Browser preview data for timing charts must include multiple completed requests plus one in-flight request, so `#frame=codex&workspace=live-sessions` exercises curve shape, latest sample footer, and live marker behavior without a Wails runtime.

docs-linhay/dev/20260523-session-distillation-codex-live-sessions-ui.md

@@ -70,6 +70,19 @@ Live sessions 的 detail 面板和筛选区本质上是工作台容器，不是
 
 这条边界适合继续放在 `gettokens-domain-engineering` 的 live sessions 小节里，不上升到 `AGENTS.md`。
 
+### 7. 请求耗时趋势只投影当前 active request
+
+请求耗时趋势图的数据应从请求记录推导，不能让视觉刷新本身成为数据来源。
+
+稳定边界：
+
+1. 已完成请求优先使用 `timing.totalDurationMs`；缺失时才回退到 `completedAt - startedAt`。
+2. 当前 active request 可以用 `nowMs - startedAt` 做实时投影，用于显示正在增长的 live 样本。
+3. 历史请求即使因为 cache 或 sidecar 残留仍带着 `streaming/reconnecting` 且没有 `completedAt`，也不能继续按 `nowMs` 投影；否则图上所有总耗时点会一起增长。
+4. 回归测试必须覆盖“stale streaming request + active request”并存时，只有 active request 增长。
+
+这类问题不要先调 CSS 或动画。先检查纯模型输出：`points[].values.totalDurationMs` 与 `points[].isLive` 是否已经错误增长；如果模型输出错，修模型，不修图表。
+
 ## 不纳入
 
 - 不新增独立 `gettokens-codex-live-sessions` skill；当前规则继续归入 `gettokens-domain-engineering` 的 Codex Live Sessions 小节。
@@ -88,5 +101,6 @@ Live sessions 的 detail 面板和筛选区本质上是工作台容器，不是
 
 - `node --test frontend/src/features/codex-live-sessions/model.test.mjs`
 - `npm --prefix frontend run typecheck`
+- `npm --prefix frontend run build`
 
 Go focused test 曾被当前工作区其他变更阻塞过：`internal/accounts/auth_file_normalize.go` 缺少 `convertSessionLikePayloadToCPA`。后续提交 live sessions 改动时，应重新确认当前工作区是否已经恢复可编译。

docs-linhay/memory/2026-05-26.md

@@ -84,3 +84,20 @@
 - 修复：`internal/sidecar/isPortFree` 改为按 sidecar 实际 `host: ""` 绑定语义探测 `:<port>`；App 启动时清理同 profile `config.yaml` 的孤儿 `cli-proxy-api`，退出时 interrupt 后等待，超时 force kill。
 - 沉淀：更新 `gettokens-ops-governance` 的 `Sidecar Process Binding` 规则，并补充 `docs-linhay/dev/20260519-macos-runtime-residency-session-distillation.md`；不升级 `AGENTS.md`，因为这是 GetTokens Wails/macOS 生命周期细则。
 - 验证：`go test ./internal/sidecar ./internal/wailsapp -run 'TestPortAvailabilityMatchesWildcardSidecarBind|TestFindOrphanedSidecarPIDsMatchesSameConfigOnly|TestStopProcessKillsWhenInterruptIsIgnored|TestGetCodexLiveSessionsSnapshot' -count=1` 通过；重新拉起当前 App 后，8317 由新 App 子进程持有，live snapshot 已返回 `projectName`，样本包含 `GetTokens` 与 `design`。
+
+## CLIProxyAPI 上游同步
+- 完成：`docs-linhay/references/CLIProxyAPI#gettokens/sidecar` 已合并 `upstream/main@167edfec`，生成并推送 merge commit `b72ac277` 到 `AxApp/CLIProxyAPI#gettokens/sidecar`。
+- 冲突：只在 `sdk/api/handlers/openai/openai_responses_websocket.go` 发生冲突；处理时保留上游 file-backed websocket timeline，同时保留 GetTokens live-session hook、pinned auth failover、wrapped status code 提取和 `coreauth.AuthAllowsWebsockets(auth)` 默认能力边界。
+- 验证：fork 内 `go test ./sdk/api/handlers/openai`、受影响包测试、`go test ./...`、`git diff --check` 均通过；父仓库 `./scripts/ensure-sidecar.sh darwin arm64` 已重建本地 sidecar，meta 记录 `commit=b72ac277`、`dirty=clean`。
+- 归档：同步记录追加到 `docs-linhay/dev/20260519-cliproxyapi-upstream-sync-session-distillation.md`；本次仍属于既有 CLIProxyAPI fork 同步流程，不升级 `AGENTS.md`。
+
+## CLIProxyAPI 内存剪裁
+- 决策：live sessions tracker 只保留最近 50 条 request，保留 30 分钟窗口但增加 `DELETE /v0/management/gettokens/live-sessions` 清理接口，避免单个长会话无限堆 request。
+- 决策：legacy `internal/usage/logger_plugin.go` 不再把 request detail 长驻内存，改为聚合常驻、明细落盘分页读取；`usage-observed` 迁移到 `usage-observed-v2.sqlite`，不再回灌旧的大 JSON payload。
+- 新增：usage attribution 增加 `GET /v0/management/gettokens/usage-attribution/details`，支持 `window / limit / offset / account_key / attribution_key`，给后续前端或调试页按页取明细。
+- 验证：`go test ./internal/usage ./internal/gettokenshooks` 通过；这轮剪裁是为了降低 sidecar RSS，不是前端展示优化，因此没有改前端 dev server 的职责边界。
+
+## Codex live sessions 请求趋势整理
+- 沉淀：请求耗时趋势图只允许当前 active request 使用 `nowMs - startedAt` 投影；历史请求保留记录值或完成时间差，避免 stale `streaming/reconnecting` 样本一起增长。
+- 已更新：`.agents/skills/gettokens-domain-engineering/SKILL.md` 的 Codex Live Sessions 小节、`docs-linhay/dev/20260523-session-distillation-codex-live-sessions-ui.md`。
+- 不升级：该规则属于 Codex runtime observability 领域建模边界，不是 repo-wide `AGENTS.md` 约束。