@modusensus/dsh-mneme 0.7.17 → 0.7.20

package/README.en.md

@@ -7,7 +7,7 @@ English | [简体中文](README.md)
 [![npm version](https://img.shields.io/npm/v/@modusensus/dsh-mneme?color=blue&label=npm)](https://www.npmjs.com/package/@modusensus/dsh-mneme)
 [![license](https://img.shields.io/badge/license-MIT-green)](LICENSE)
 [![Awesome](https://awesome-dsh-plugin.com/badge.svg)](https://github.com/awesome-dsh-plugin/awesome-dsh-plugin)
-[![tests](https://img.shields.io/badge/tests-664%20passed-success)](https://github.com/modusensus/dsh-mneme)
+[![tests](https://img.shields.io/badge/tests-685%20passed-success)](https://github.com/modusensus/dsh-mneme)
 
 > A cross-session memory plugin for DeepSeek Harness: it lets the Agent remember you, remember your projects, and organize memories automatically. **Mneme** (Μνήμη) — named after Mnemosyne, the Greek goddess of memory who presides over memory and dreams, just as autoDream consolidates memories in the background.
 
@@ -162,10 +162,12 @@ Every **background LLM call** (autoDream consolidation + summary, autoSummarize
 
 ## 🆕 Recent Release Highlights
 
-> ⚠️ **Archival note**: pre-v0.7.12 rows below record experimental features later removed in the v0.7.12 rewrite (heat model, Wiki-Link, tag system/directory/tag-boost, user/fact layered types, prefix-id resolution, /stats and /directory endpoints). They are version history only — **not current capability**. Current features are what this README's body and the [config table](#-configuration) describe.
+> ⚠️ **Archival note**: pre-v0.7.12 rows below record experimental features later removed in the **v0.7.11** rewrite (Wiki-Link, tag system/directory/tag-boost, user/fact layered types, prefix-id resolution, /stats and /directory endpoints). v0.7.11 and v0.7.12 shipped the same day; historical docs credited v0.7.12, corrected here to v0.7.11. Unrecorded removals in the same rewrite: session lifecycle (`session_disposed_at` soft-hide), provenance birth-session tracking (`session_id`), and the decision-field normalizer (`normalizeDecisions`, superseded by strict prompt schema + `extractJsonArray`). They are version history only — **not current capability**. (The heat model was fully restored in v0.7.20 from v0.7.10, see below.) Current features are what this README's body and the [config table](#-configuration) describe.
 
 | Version | Highlights |
 |------|------|
+| **v0.7.20** | Heat model restored (issue #87): v0.7.0 self-evolving memory back (`src/heat.js` power-law decay `H=1/(1+λΔt)^α` + per-type half-lives), sleep demotion dual protection (cold time-window + heat<0.05 + importance<5), touchRecalled gating back on `heatEnabled`, entity heat projection (ego node heat → size/opacity), recall_runs recording on by default; **default OFF** (matches v0.7.12 behavior out of the box) + feature_flags whitelist rollback switch + lightMode linkage + sleep demotion audit counts exposed (workbench can show "N demoted") + phase-two frontend (/list heat projection, HeatBadge three-tier badge, self-gated status heat-distribution card, order=heat page-local sort, all self-gated); better-sidebar fix (issue #88: soft integration moved to an inner dynamic sub-plugin, no more startup failure without bs); 685 tests green |
+| **v0.7.18** | Ecosystem step one + query convergence: better-sidebar soft integration (inject declaration + optional peer `dsh-better-sidebar` + registerTab reusing the four views, safe skip when absent; narrow-container `@container` adaptation) + `/list?deposited=only` view (receipt_chain ∪ source=dream) + deposited/archived filter chips in the library + status-page dashboard (server totals + view-all jumps) + drawer restore for archived memories; 667 tests green |
 | **v0.7.17** | Panel polish: sidebar entry continuously aligned with the host (MutationObserver syncs the New-Session class + `width:100%` + native centering restored, resilient to async skin rewrites) + importance rendered as Lucide star glyphs (`ImportanceStars` filled/hollow component replacing text ★) + toolbar dropdown stacking fix (z-index moved to the container; export/import menu no longer painted under the sticky month header); 664 tests green |
 | **v0.7.16** | Fixed autoDream empty-body failures on thinking models (`no json array in llm output`): restored config-first routing (settings "consolidation model" wins, Issue #25) + reasoningEffort auto-retry without effort on rejection + honest `llm_audit` error on parse failure; backfilled API-route tests (/delete, /entities, /external-api) + lib runtime smoke; 662 tests green |
 | **v0.7.15** | Desktop adaptation: library panel redesign + 30-key feature-flag UI (features API) + status dashboard + import/export (mirror-isomorphic md golden loop) + token masking by default; 645 tests green |
@@ -210,7 +212,7 @@ Every **background LLM call** (autoDream consolidation + summary, autoSummarize
 | v0.7.1–0.7.8 | ✅ Done | Issue fixes + graph backfill | tags↔entity_attrs bridge, inline-confirm delete, sidebar trigger toggle, brace escaping, user/fact layered types + stats endpoint (later removed), prefix-id resolution (later removed), sleep batch entity extraction, snapshotEvents() DSH compat; 764→812 tests green |
 | v0.7.9 | ✅ Done | lib-sync gate | Issue #65: src-only shim silently killed the shipped lib — synced lib + pre-publish src↔lib consistency check + lib smoke tests; 815 tests green |
 | v0.7.10–0.7.12 | ✅ Done | Panel redesign + near-rewrite | UX upgrades (color dots, pan/zoom, pagination, global search), then near-rewrite: inlined panel → pure HTTP API + standalone CLI + lightMode; TYPES narrowed 8→6 |
-| v0.7.13–0.7.17 | ✅ Done | Distillation, security, desktop, dream fix, panel polish | codingRetrospect + 429 governor; private reasoning blocks dropped (CWE-200); desktop panel redesign + feature flags; autoDream thinking-model empty-body fix + backfilled tests; sidebar entry alignment + star glyphs + menu stacking polish; 664 tests green |
+| v0.7.13–0.7.20 | ✅ Done | Distillation, security, desktop, dream fix, panel polish, ecosystem, heat restore | codingRetrospect + 429 governor; private reasoning blocks dropped (CWE-200); desktop panel redesign + feature flags; autoDream thinking-model empty-body fix + backfilled tests; sidebar entry alignment + star glyphs + menu stacking polish; better-sidebar soft integration (inner dynamic sub-plugin fix, issue #88) + deposited/archived library views; heat model restored (issue #87) with phase-two frontend + archival audit (issue #87); 685 tests green |
 | **v0.8.0** | 🚧 Planned (late Sep) | Graph enhancement | Interest-drift visualization + scope isolation (issue #17) + cross-workspace sharing |
 
 > All new capabilities ship as **toggleable features** (enabled/disabled via configuration), conservatively on by default and never breaking existing behavior. The `failure_memories` table and the autoDream decision engine have already paved the way for future reflective growth.
@@ -316,6 +318,7 @@ It works out of the box with the defaults. To adjust, override in `~/.dsh/profil
 | `evalPersistTestResults` | `false` | Retrieval evaluation persistence (v0.4.5, opt-in, off by default): when enabled, `evaluateRetrieval` writes precision/recall/mrr snapshots into `recall_evals`; when off it only returns them to the caller without persisting. Production `searchMemories` auditing always goes through `recall_runs`, unconditionally never touching `recall_evals` |
 | `autoReindexOnBoot` | `true` | When existing memories lack embeddings and vectors are configured, a delayed, rate-limited background backfill rebuild starts after boot (set to `false` for manual rebuild only) |
 | `hybridInject` | `true` | Semantic-recall-first injection (v0.4.6, Bug4): when `injectCandidates` receives a non-empty query, it first recalls candidates via the vector index semantically, then fills in/deduplicates with rule-based filtering; empty query / no vectors falls back to the old logic |
+| `heatEnabled` | `false` | Heat-model master switch (v0.7.0 / restored in v0.7.20, **off by default** — v0.7.12+ users expect no-heat behavior): when on, provides heat fields / sleep heat-combined demotion protection / frontend heat projection without changing recall ranking; when off, skips all heat computation and touch, and sleep demotion reverts to pure time-tiering. Also on the feature_flags whitelist (panel-toggleable rollback switch); light-mode preset forces it off |
 | `memoryQualityFilter` | `{enabled:true, archiveThreshold:30, degradeThreshold:60, minContentLength:10}` | Memory quality filter (v0.4.6, on by default): heuristic 0-100 scoring before write; meta-memory vocabulary/self-reference/overly short/duplicate/near-duplicate content deduct points; ≥60 stored normally, 30-60 down-weighted (injection ranked by importance×quality/100), <30 archived and flagged `low_quality` (still recallable via explicit search, never auto-injected) |
 | `llmAudit` | `{enabled:true, retentionDays:90}` | LLM usage audit (v0.4.6, on by default): every background LLM call (autoDream/autoSummarize) writes `llm_audit_logs` (tokens/duration/status/source); failures are recorded as error without blocking; read-only APIs `/api/dsh-mneme/semantic/llm-audit` + `/llm-audit/stats` |
 

package/README.md

@@ -5,7 +5,7 @@
 [![npm version](https://img.shields.io/npm/v/@modusensus/dsh-mneme?color=blue&label=npm)](https://www.npmjs.com/package/@modusensus/dsh-mneme)
 [![license](https://img.shields.io/badge/license-MIT-green)](LICENSE)
 [![Awesome](https://awesome-dsh-plugin.com/badge.svg)](https://github.com/awesome-dsh-plugin/awesome-dsh-plugin)
-[![tests](https://img.shields.io/badge/tests-664%20passed-success)](https://github.com/modusensus/dsh-mneme)
+[![tests](https://img.shields.io/badge/tests-685%20passed-success)](https://github.com/modusensus/dsh-mneme)
 [![CI](https://img.shields.io/github/actions/workflow/status/modusensus/dsh-mneme/ci.yml)](https://github.com/modusensus/dsh-mneme/actions)
 [![node](https://img.shields.io/badge/node-24%2B-blue)](https://nodejs.org)
 [![npm downloads](https://img.shields.io/npm/dm/@modusensus/dsh-mneme?color=blue&label=downloads)](https://www.npmjs.com/package/@modusensus/dsh-mneme)
@@ -34,7 +34,7 @@ dsh web
 - 不是向量数据库——语义搜索是可选增强，默认零额外依赖
 - 不替代会话日志——它存的是「值得跨会话记住的」精炼知识
 - 不改变模型本身——进化的是记忆库与每次注入的上下文
-- 删对话 ≠ 删记忆——开启会话生命周期后，删除会话只是把该会话出生的记忆**软隐藏**（可恢复），数据不丢
+- 删对话 ≠ 删记忆——记忆是跨会话的全局知识（v0.7.11 起不再与会话绑定），删除会话不影响已存入的记忆
 
 ## ✨ 功能
 
@@ -42,7 +42,7 @@ dsh web
 
 - **SQLite 主存储**：`~/.dsh/memory/memory.db`，`node:sqlite` 内置，零原生依赖
 - **Markdown 镜像**：`preferences.md` / `projects.md` / `decisions.md` / `history.md` / `summary.md`，人类可读、可手工编辑（**人工修改优先**合并回库）
-- **9 种记忆类型**：`preference` 偏好 / `project` 项目 / `decision` 决策 / `history` 历史 / `summary` 会话总览 / `pattern` 模式 + 编码记忆三型 `rejected_solution` 被否方案 / `pitfall` 踩坑 / `constraint` 约束（v0.7.13 起，`codingRetrospect` 默认关；v0.7.12 重写已收窄删除 user/fact 两型）
+- **9 种记忆类型**：`preference` 偏好 / `project` 项目 / `decision` 决策 / `history` 历史 / `summary` 会话总览 / `pattern` 模式 + 编码记忆三型 `rejected_solution` 被否方案 / `pitfall` 踩坑 / `constraint` 约束（v0.7.13 起，`codingRetrospect` 默认关；v0.7.11 近重写已收窄删除 user/fact 两型）
 - **镜像同步状态机（v0.3.6+）**：mirror 与主库强一致，用 `generation`（期望轮次）/ `applied_generation`（已应用轮次）建模同步债务
   - 业务写操作在**自身事务内原子递增** desired generation——崩溃在 COMMIT 后、渲染前，重启也能凭 durable 债务恢复，绝不静默跳过（v0.3.8）
   - `generation` 用 SQLite 原子语句递增，多进程并发零丢失；带 `CHECK` 上界，负数/溢出拒绝
@@ -107,15 +107,6 @@ dsh web
 
 > 配置详见 `docs/SLEEP.md`；迁移说明见 `docs/MIGRATION.md`。
 
-### 会话生命周期：把对话当存档点 💾（v0.6.0，opt-in）
-
-**默认关闭**（`sessionLifecycleEnabled: false`，保持旧行为）。开启后，会话被删除/销毁（DSH `session/disposed` 事件）时，自动把该会话内出生（`session_id` 溯源）的记忆**软隐藏**——不再出现在检索/注入/列表/整理，但**不删除**，随时可恢复：
-
-- **与 `archived` 正交**：`archived` 是用户/AI 主动"长期保留但安静"，`session_disposed_at` 是会话删除被动隔离，两者互不覆盖。恢复会话绝不复活你手动归档的记忆
-- **全局记忆免疫**：存量无 `session_id` 的记忆视为全局，永不参与会话清理
-- **幂等 + 熔断**：dispose/restore 状态守卫幂等（重复调用 no-op）；事件回调内部异常 catch 住，不抛进 DSH 会话清理流程
-- **恢复**：整会话 `service.restoreBySession(sessionId)` 一键还原；`service.listBySession(sessionId, { includeDisposed: true })` 可查看当前隐藏了哪些（DTO 带 `disposed` 标记）
-
 官方设置面板 → 「记忆库设置」→「记忆」标签：按类型浏览、全文搜索；启用向量搜索后可用「语义」切换做向量召回。
 
 ### 用户设置（画像 / 规则）与自定义指令 ⚙️
@@ -212,10 +203,12 @@ v0.3.0 起新增**记忆基因**层：从记忆里抽取**命名实体**、**带
 
 ## 🆕 最近版本亮点
 
-> ⚠️ **历史存档提示**：下方及路线图表中 v0.7.12 之前的早期条目记录的实验功能（heat 热度模型、Wiki-Link、tag 系统/目录/tag 加权、user/fact 分层类型、前缀 id 解析、/stats 与 /directory 端点等）已在 v0.7.12 近重写时移除，仅作版本历史存档，**不代表当前能力**。当前特性以本 README 正文与 [配置表](#-配置) 为准。
+> ⚠️ **历史存档提示**：下方及路线图表中 v0.7.12 之前的早期条目记录的实验功能（Wiki-Link、tag 系统/目录/tag 加权、user/fact 分层类型、前缀 id 解析、/stats 与 /directory 端点等）已在 **v0.7.11** 近重写时移除（v0.7.11 与 v0.7.12 同日发布，历史文档多标 v0.7.12，此处勘误归一到 v0.7.11），仅作版本历史存档，**不代表当前能力**。同期未记载的移除另有：会话生命周期（`session_disposed_at` 软隐藏）、provenance 出生会话溯源（`session_id`）、决策字段归一化（normalizeDecisions，已由提示词硬规范取代）。（heat 热度模型已于 v0.7.20 从 v0.7.10 完整找回回归，见下方 [v0.7.20](#recent-version-highlights)。）当前特性以本 README 正文与 [配置表](#-配置) 为准。
 
 | 版本 | 亮点 |
 |------|------|
+| **v0.7.20** | heat 热度模型回归（issue #87）：找回 v0.7.0 自进化记忆（`src/heat.js` 幂律衰减 `H=1/(1+λΔt)^α` + per-type 差异化半衰期）、sleep 降级热联合双保护（时间窗冷 + heat<0.05 + importance<5）、touchRecalled 门控改回 `heatEnabled`、实体热投影（ego 节点 heat → 前端大小/明暗）、recall_runs 默认记录；**默认关**（默认=与 v0.7.12 行为一致）+ feature_flags 白名单回滚开关 + lightMode 联动 + sleep 降级审计计数暴露（工作动态可展示"降级 N 条"）+ 阶段二前端（/list heat 投影、HeatBadge 三档徽章、状态页热度分布卡、order=heat 页内排序，全部自门控）；better-sidebar 修复（issue #88：软集成改内层动态子插件，无 bs 环境不再启动失败）；685 测试全绿 |
+| **v0.7.18** | 生态第一步 + 查询收敛：better-sidebar 软集成（inject 声明 + optional peer `dsh-better-sidebar` + registerTab 复用四视图，未装安全跳过；窄容器 `@container` 适配）+ `/list?deposited=only` 沉淀视图（receipt_chain ∪ source=dream）+ 记忆库沉淀/已归档筛选 chip + 状态页仪表盘化（统计 + 查看全部跳转预置筛选）+ 抽屉归档记忆「恢复」；667 测试全绿 |
 | **v0.7.17** | 面板体验细化：侧边栏入口持续对齐宿主（MutationObserver 同步「新会话」类名 + `width:100%` + 交还原生居中，皮肤异步改写不再失配）+ 重要性星级换 Lucide 星形（`ImportanceStars` 实心/空心组件，替换文本 ★）+ 工具栏下拉层级修复（z-index 提到容器，导出/导入菜单不再被吸顶月份头遮挡）；664 测试全绿 |
 | **v0.7.16** | 修复 autoDream 在 thinking 模型上空体 failed（`no json array in llm output`）：恢复 config-first 路由（设置面板「巩固模型」生效，Issue #25）+ reasoningEffort 被拒自动去掉重试一次 + 解析失败如实记 llm_audit error 并带原始输出头日志；补测 API 路由空白（/delete、/entities、/external-api）+ lib 运行时冒烟（src↔lib 一致性）；662 测试全绿 |
 | **v0.7.15** | 桌面端适配：记忆库面板重设计（撤对话 tab → 居中非全屏 sheet、卡片网格/时间线双视图、右侧详情抽屉、编辑/归档/关联实体）+ 侧边栏入口上移工作区上方（借用宿主原生类名对齐、收起态零位移）+ 功能开关 30 键上 UI（features API，巩固模型与 embedding 提供方可配）+ 状态页工作台（巩固卡/工作动态流/沉淀记忆/归档恢复）+ 导入导出（镜像同构 md 黄金闭环）+ Token 面板默认遮蔽；645 测试全绿 |
@@ -346,8 +339,6 @@ dsh web
     dreamThresholdCount: 10
     dreamThresholdChars: 5000
     dreamDelayMs: 2000
-    # 会话生命周期（v0.6.0，默认关）：会话被删时软隐藏其记忆，可恢复
-    # sessionLifecycleEnabled: false
 ```
 
 ## ⚙️ 配置
@@ -403,6 +394,12 @@ dsh web
 | `bm25SearchEnabled` | `true` | BM25 稀疏第三路召回（v0.5.0）：ASCII 词元 + CJK bigram，IDF 加权，散词/ID/代码片段查询不再依赖子串命中 |
 | `adaptiveThresholdEnabled` | `true` | 自适应相似度阈值（v0.5.0）：按查询形态动态截断（前缀 0.5 / 短查询 0.7 / 长查询 0.6 / 头部分差大放宽 0.5），显式传 `threshold` 走旧行为 |
 | `hotMemoryEnabled` | `true` | 会话级短期热记忆总开关（v0.5.0）：关闭后热记忆块不再注入（长期召回不受影响） |
+| `heatEnabled` | `false` | 热度模型总开关（v0.7.0 / v0.7.20 回归，**默认关**——v0.7.12 起用户已习惯无 heat 行为）：开启后提供热度字段 / sleep 热联合降级保护 / 前端热度投影，不改变召回排序；关闭则跳过 heat 计算与热度触达，sleep 降级退回纯时间分层。也走 feature_flags 白名单（面板可启停=回滚开关），lightMode 预设强制关 |
+| `heatGlobalAlpha` | `1.2` | 幂律形状参数 α（`H=1/(1+λΔt)^α`），越大衰减越快 |
+| `heatTypeDecay` | 内置 TYPE_DECAY | per-type 衰减因子 λ；λ=0 的类型免疫（preference/pattern/summary 热度恒 1.0，sleep 永不降级） |
+| `sleepHeatThreshold` | `0.05` | sleep 降级联合判定热度下限：heat<该值 **且** importance<5 才允许降级 |
+| `recallRecordDefault` | `true` | recall_runs 记录默认开（显式传 `recordRecall:false` 的调用方不受影响） |
+| `recallRetentionDays` | `90` | recall_runs 滚动清理保留天数 |
 | `hotMemoryRounds` | `5` | 会话级短期热记忆轮次（v0.5.0）：最近 N 轮对话滚动注入，从会话事件日志无状态重建、不落库 |
 | `hotMemoryMaxTokens` | `2000` | 热记忆 token 预算（v0.5.0，200-32000），超出滚动截断 |
 | `selectiveInjectEnabled` | `true` | 选择性注入（v0.5.0）：query 向量可用时注入候选按主题相似度重排，替代固定规则序 |
@@ -410,8 +407,6 @@ dsh web
 | `searchSemanticDedupThreshold` | `0.95` | 语义去重相似度阈值（v0.5.0，默认 0.95，范围 0.5-1.0）：`searchSemanticDedup=true` 时生效，调整可防小模型误折叠 |
 | `memoryQualityFilter` | `{enabled:true, archiveThreshold:30, degradeThreshold:60, minContentLength:10}` | 记忆质量过滤（v0.4.6，默认开）：写库前启发式打分 0-100，元记忆词汇/自指/过短/重复/近似重复扣分；≥60 正常存储，30-60 降权（注入排序按 importance×quality/100），<30 归档标记 `low_quality`（显式搜索仍可召回，永不自动注入） |
 | `llmAudit` | `{enabled:true, retentionDays:90}` | LLM 消耗审计（v0.4.6，默认开）：每次后台 LLM 调用（autoDream/autoSummarize）写 `llm_audit_logs`（tokens/duration/status/source）；失败记 error 不阻塞；只读 API `/api/dsh-mneme/semantic/llm-audit` + `/llm-audit/stats` |
-| `sessionLifecycleEnabled` | `false` | 会话生命周期（v0.6.0，默认关）：开启后会话被删除/销毁时自动把该会话出生的记忆软隐藏（`session_disposed_at`，与 `archived` 正交、可恢复）；存量无 `session_id` 的记忆永不参与清理 |
-| `showSidebarTrigger` | `true` | 侧边栏底部（左下角）记忆入口按钮（issue #38，默认开）：与其他插件（如 dsh-cost-meter）抢占同一 footer slot 导致 UI 冲突时可关掉；仅隐藏按钮，记忆库仍可通过顶部「记忆库」标签访问。Web 面板「设置」里有对应开关 |
 
 > 🔐 **API 安全**：DSH 无内置鉴权且默认仅监听 `127.0.0.1`。插件 API 默认开放（便于 Web 面板即装即用）。如需防护（如局域网暴露），在配置中设置 `apiToken`：写操作（画像/规则/命令）与密钥端点（`vector-config`、`vector-reindex`）需携带 `Authorization: Bearer <token>`（前端设置面板可填入同一 token），只读的 `list` / `search` / `semantic` 保持开放。`/api/dsh-mneme/vector-config` 返回的 `apiKey` 已掩码（`sk-***…`），存储仍保留明文供调用；前端回传空或掩码值表示"不改 key"。
 

package/cordis.patch.yml

@@ -13,4 +13,9 @@
         dreamThresholdCount: 10
         dreamThresholdChars: 5000
         dreamDelayMs: 2000
+        # 巩固默认压低推理：思考型模型（如方舟 v4 系）会把 dreamMaxTokens
+        # 预算花在推理上、正文空体 → "no json array in llm output"。low 让
+        # provider 封顶推理；拒绝 effort 字段的 provider 由 runAuditedLlm
+        # 自动去掉重试，行为安全（v0.7.16 机制）。
+        dreamReasoningEffort: low
         entityExtractionEnabled: true

package/lib/api.js

@@ -3,6 +3,7 @@ import { readFileSync } from "node:fs";
 import { timingSafeEqual, randomBytes } from "node:crypto";
 import { FEATURE_FLAG_SPEC } from "./settings.js";
 import { TYPE_FILE, renderMirrorText, parseHumanEdits } from "./mirror.js";
+import { computeHeat } from "./heat.js";
 
 // headers：少数端点（/export 附件下载）需要追加 Content-Disposition 等响应头。
 function sendJson(res, status, payload, headers = {}) {
@@ -178,18 +179,26 @@ export function createApi(ctx, service, settings, commands, embedder, semantic =
         // archived=only：只看归档（状态页的归档列表用）。归档行不进默认列表，
         // 所以这是独立的视图开关，而不是 includeArchived 的混看模式。
         const onlyArchived = url.searchParams.get("archived") === "only";
-        const rows = service.list({ type, limit, offset, order, minImportance, source, updatedFrom, updatedTo, onlyArchived });
+        // deposited=only：只看 autoDream 巩固过的记忆（receipt_chain 的
+        // merge/update verdict ∪ source=dream 直写）——状态页「查看全部」
+        // 与记忆库的「沉淀」筛选 chip 共用这个视图。
+        const depositedOnly = url.searchParams.get("deposited") === "only";
+        const rows = service.list({ type, limit, offset, order, minImportance, source, updatedFrom, updatedTo, onlyArchived, depositedOnly });
         // 面板行在 wire DTO 之上补 archived/quality_score——模型工具的输出
         // schema 严格复用 toApiList，扩展只发生在 HTTP 层。
+        // heat 投影（阶段二前端数据源）：仅 heatEnabled=true 时下发逐条热度
+        // （heat.js 纯函数，λ=0 免疫类型恒 1.0）；字段缺省时前端徽章自动隐藏。
+        const heatOn = config?.heatEnabled === true;
         const items = service.toApiList(rows).map((m, i) => ({
           ...m,
           archived: rows[i].archived === true || rows[i].archived === 1,
-          quality_score: rows[i].quality_score ?? null
+          quality_score: rows[i].quality_score ?? null,
+          ...(heatOn ? { heat: computeHeat(rows[i], Date.now(), config ?? {}) } : {})
         }));
         // Total honors the same filters as the rows, or the pager's
         // has-more math breaks whenever minImportance/source/updated-at
         // bounds are active.
-        sendJson(res, 200, { items, total: service.count(type, { minImportance, source, updatedFrom, updatedTo, onlyArchived }) });
+        sendJson(res, 200, { items, total: service.count(type, { minImportance, source, updatedFrom, updatedTo, onlyArchived, depositedOnly }) });
       } catch {
         sendJson(res, 500, { error: "internal" });
       }
@@ -586,7 +595,10 @@ export function createApi(ctx, service, settings, commands, embedder, semantic =
             name: n.name,
             type: n.type ?? null,
             mention_count: n.mention_count ?? 1,
-            distance: n.distance
+            distance: n.distance,
+            // v0.7.0 实体热投影：实体热 = 关联记忆 heat 聚合（max），前端据此
+            // 缩放节点大小/明暗。heatEnabled=false 时 entityHeat 返回 null。
+            heat: service.entityHeat?.(n.id) ?? null
           })),
           edges: [...edgeMap.values()].map((e) => ({
             id: e.id,
@@ -775,7 +787,13 @@ export function createApi(ctx, service, settings, commands, embedder, semantic =
           status: r.status,
           provider: r.provider ?? null,
           model: r.model ?? null,
-          error: r.error ?? null
+          error: r.error ?? null,
+          run_type: r.run_type ?? "auto",
+          // sleep 审计：heat/时间分层降级决策计数（工作动态可展示"降级 N 条"）。
+          // 数据来自 runSleep 写入的 decisions.demotion（{demoted,archived} 数组）。
+          demotion: r.decisions?.demotion
+            ? { demoted: (r.decisions.demotion.demoted ?? []).length, archived: (r.decisions.demotion.archived ?? []).length }
+            : null
         }));
         const pendingConflicts = service.countConflictPending?.() ?? 0;
         const ids = new Set();