feat: integrate session lifecycle protocol into chat hooks and sidebar

CHANGELOG.md

@@ -1,5 +1,45 @@
 # Changelog
 
+## Unreleased
+
+### Public Upstream Sync - 2026-04-11
+
+#### Included in this upstream-safe branch
+- Chat composer queue and steer workflow:
+  - Added durable queued-turn data structures and queue reconciliation helpers.
+  - Added steer promotion, paused queue handling, queued-turn resume, and queue cleanup after turn settlement.
+  - Added queue/session-scope tests for `codexQueue`, `sessionLoadGuards`, `sessionSnapshotCache`, and `sessionScope`.
+- Session stability and protocol hardening:
+  - Added explicit WebSocket lifecycle protocol messages: `session-accepted`, `session-busy`, `session-state-changed`.
+  - Added project-scoped event enrichment for session lifecycle payloads to avoid cross-session/cross-project ambiguity.
+  - Added lifecycle projection from provider completion/error messages into normalized session-state events.
+  - Added session-created `projectName` metadata for Claude/Cursor/Gemini session initialization paths.
+  - Removed unused `projectPath` from `session-accepted` payloads; downstream should rely on `projectName` + scoped identifiers.
+- Nano chain compatibility:
+  - UI provider/model selection no longer surfaces Nano by default in the upstream-safe branch, while server-side `nano-command` handling remains for compatibility.
+  - Preserved Nano command path and active session reporting in WebSocket session status flows.
+  - Ensured session lifecycle protocol is emitted consistently for Nano just like other providers.
+
+#### Migration notes
+- **Message cache key change**: Session message cache keys are now scoped by provider (`chat_messages_{project}_{provider}_{session}`). Existing localStorage entries keyed the old way (`chat_messages_{project}_{session}`) will not be read by default. To migrate, set `allowLegacyFallback: true` when calling `getSessionMessageCacheKeys()`. Users upgrading from builds prior to this sync may see empty transcript history on first load for previously-active sessions; the data is still in localStorage and can be recovered by enabling the legacy fallback or manually re-keying the entries.
+
+#### Explicitly excluded (local/private only, not part of upstream PR)
+- Codex-only product strategy and provider lock-in controls.
+- External auth / license refresh-heartbeat-offline-grace stack.
+- Project root hard-cut and Codex session backfill private policy behavior.
+- LingZhi/LingzhiLab branding replacements and private demo/link route changes.
+
+#### Cross-platform notes
+- No platform-specific server behavior was hardcoded for this sync.
+- Session protocol additions are transport-level and provider-agnostic; Windows-specific stability paths do not alter macOS behavior.
+
+#### Validation
+- Passed: `node --check server/index.js`
+- Passed: `node --check server/claude-sdk.js`
+- Passed: `node --check server/cursor-cli.js`
+- Passed: `node --check server/gemini-cli.js`
+- Not runnable in current environment (missing local dev dependencies): `vitest`, `tsc`
+
 ## Dr. Claw v1.1.1 - 2026-03-30
 
 ### Highlights

README.md

@@ -644,7 +644,7 @@ When you first open Dr. Claw you will see the **Projects** sidebar. You have two
 - **Open an existing project** — Dr. Claw auto-discovers registered projects and linked sessions from Claude Code, Codex, and Gemini.
 - **Create a new project** — Click the **"+"** button, choose a directory on your machine, and Dr. Claw will set up the workspace: agent folders such as `.claude/`, `.agents/`, `.gemini/`, standard workspace metadata, linked `skills/` directories, preset research dirs (`Survey/references`, `Survey/reports`, `Ideation/ideas`, `Ideation/references`, `Experiment/code_references`, `Experiment/datasets`, `Experiment/core_code`, `Experiment/analysis`, `Publication/paper`, `Promotion/homepage`, `Promotion/slides`, `Promotion/audio`, `Promotion/video`), and **instance.json** at the project root with absolute paths for those directories. Cursor agent support is coming soon.
 
-> **Default project storage path:** New projects are stored under `~/dr-claw` by default. You can change this in **Settings → Appearance → Default Project Path**, or set the `WORKSPACES_ROOT` environment variable. The setting is persisted in `~/.claude/project-config.json`.
+> **Default project storage path:** New projects are stored under `~/dr-claw` by default. You can change this in **Settings → Appearance → Default Project Path**, or set the `WORKSPACES_ROOT` environment variable. The setting is persisted in `~/.dr-claw/project-config.json` (with automatic one-time migration from `~/.claude/project-config.json`).
 
 </details>
 

README.zh-CN.md

@@ -640,7 +640,7 @@ Dr. Claw 的核心功能是 **Research Lab**。
 - **打开已有项目** — Dr. Claw 会自动发现已注册项目，以及来自 Claude Code、Codex、Gemini 的关联会话。
 - **创建新项目** — 点击 **"+"** 按钮，选择本机的一个目录，Dr. Claw 会创建：`.claude/`、`.agents/`、`.gemini/` 等 Agent 目录、标准工作区元数据、链接的 `skills/` 目录、预设研究目录（`Survey/references`、`Survey/reports`、`Ideation/ideas`、`Ideation/references`、`Experiment/code_references`、`Experiment/datasets`、`Experiment/core_code`、`Experiment/analysis`、`Publication/paper`、`Promotion/homepage`、`Promotion/slides`、`Promotion/audio`、`Promotion/video`），以及项目根目录下的 **instance.json**（上述目录的绝对路径写入其中）。Cursor Agent 支持即将推出。
 
-> **默认项目存储路径：** 新项目默认存储在 `~/dr-claw` 目录下。可在 **Settings → Appearance → Default Project Path** 中修改，也可通过环境变量 `WORKSPACES_ROOT` 设置。该配置持久化在 `~/.claude/project-config.json` 中。
+> **默认项目存储路径：** 新项目默认存储在 `~/dr-claw` 目录下。可在 **Settings → Appearance → Default Project Path** 中修改，也可通过环境变量 `WORKSPACES_ROOT` 设置。该配置持久化在 `~/.dr-claw/project-config.json` 中（会自动一次性迁移 `~/.claude/project-config.json`）。
 
 </details>
 

src/components/chat/hooks/__tests__/useChatSessionState.test.ts

@@ -0,0 +1,24 @@
+import { describe, expect, it } from 'vitest';
+
+import {
+  hasPendingOptimisticSessionState,
+  hasTemporaryProcessingSessionKeys,
+} from '../useChatSessionState';
+
+describe('useChatSessionState temporary session helpers', () => {
+  it('treats temp sessions as pending optimistic sessions', () => {
+    expect(hasPendingOptimisticSessionState(null, 'new-session-123')).toBe(true);
+    expect(hasPendingOptimisticSessionState(null, 'temp-123')).toBe(true);
+    expect(hasPendingOptimisticSessionState({ sessionId: null, startedAt: Date.now() }, null)).toBe(true);
+    expect(hasPendingOptimisticSessionState(null, 'sess-123')).toBe(false);
+  });
+
+  it('detects temporary processing keys for both raw and scoped temp ids', () => {
+    expect(hasTemporaryProcessingSessionKeys(new Set(['new-session-1']))).toBe(true);
+    expect(hasTemporaryProcessingSessionKeys(new Set(['temp-1']))).toBe(true);
+    expect(hasTemporaryProcessingSessionKeys(new Set(['proj-a::codex::new-session-2']))).toBe(true);
+    expect(hasTemporaryProcessingSessionKeys(new Set(['proj-a::codex::temp-2']))).toBe(true);
+    expect(hasTemporaryProcessingSessionKeys(new Set(['proj-a::codex::sess-2']))).toBe(false);
+    expect(hasTemporaryProcessingSessionKeys(new Set(['sess-2']))).toBe(false);
+  });
+});