feat(agent-workspace): signal stable history streaks

Author: Jacobinwwey

docs/diataxis/en/explanation/agent-conversation-focus-mode-plan.md

@@ -472,10 +472,11 @@ It is conversation-product hardening on top of the shipped shell.
 `M8.34` now explains why the primary follow-up stayed ahead of the surfaced secondary alternative, so expanded history items no longer leave the primary-vs-secondary choice implicit.
 `M8.35` now explains recommendation drift across adjacent history events, so the current-atom history surface no longer requires manual cross-card comparison for that change.
 `M8.36` now explains recommendation stability across adjacent history events, so the current-atom history surface also shows when a newer event kept the same next step in place.
+`M8.37` now adds stability signals on top of that explanation surface, so repeated hold streaks are readable from badges and confidence copy without reading the full stability paragraph.
 
 Priority order:
 
-1. deepen personalization on top of the shipped primary/secondary/drift/stability explanation surface,
+1. deepen personalization on top of the shipped primary/secondary/drift/stability explanation surface and stability signals,
 2. expand broader operator rollups only when they clearly compound the same current-atom continuity loop,
 3. keep markdown-reader improvements secondary unless they clearly compound the same study loop.
 

docs/diataxis/en/explanation/development-progress-dashboard.md

@@ -6,7 +6,7 @@ It explicitly separates branch-level implementation status from mainline integra
 ## Scope
 
 - Focus area: local-first knowledge mastery platform (ingest, retrieval, learning path, tutor, memory, governance).
-- Time window: `v1.7.0` through the 2026-04-21 foundation-integration checkpoint, the 2026-04-21 UTC CI review cycle, and the post-`M8.36` reconciliation pass.
+- Time window: `v1.7.0` through the 2026-04-21 foundation-integration checkpoint, the 2026-04-21 UTC CI review cycle, and the post-`M8.37` reconciliation pass.
 - Evidence rule: every progress claim must map to:
   - contract layer (`src/learning/api.ts`, `src/learning/types.ts`),
   - route/runtime layer (`src/server.ts`),
@@ -18,12 +18,12 @@ Current top priority is not feature-count expansion.
 Current top priority is **conversation-product hardening on top of the already-green backend baseline**:
 
 - keep backend sufficiency as a standing decision gate rather than an active execution lane,
-- keep closing the correction/product/operator gap exposed after `M8.20-M8.36`,
-- treat `M8.23` as the operator-visibility baseline for conversation-memory/correction actions, `M8.24` as the active-atom action-surface baseline, `M8.25` as the active-atom pane-state continuity baseline, `M8.26` as the active-atom study-loop summary baseline, `M8.27` as the active-atom recent-result continuity baseline, `M8.28` as the active-atom result-preview baseline, `M8.29` as the active-atom history-drill-down baseline, `M8.30` as the active-atom history follow-up baseline, `M8.31` as the active-atom history freshness-ranking baseline, `M8.32` as the active-atom history rationale/alternative baseline, `M8.33` as the active-atom history confidence-signal baseline, `M8.34` as the primary-vs-secondary tradeoff explanation baseline, `M8.35` as the adjacent-history drift explanation baseline, and `M8.36` as the adjacent-history stability explanation baseline,
+- keep closing the correction/product/operator gap exposed after `M8.20-M8.37`,
+- treat `M8.23` as the operator-visibility baseline for conversation-memory/correction actions, `M8.24` as the active-atom action-surface baseline, `M8.25` as the active-atom pane-state continuity baseline, `M8.26` as the active-atom study-loop summary baseline, `M8.27` as the active-atom recent-result continuity baseline, `M8.28` as the active-atom result-preview baseline, `M8.29` as the active-atom history-drill-down baseline, `M8.30` as the active-atom history follow-up baseline, `M8.31` as the active-atom history freshness-ranking baseline, `M8.32` as the active-atom history rationale/alternative baseline, `M8.33` as the active-atom history confidence-signal baseline, `M8.34` as the primary-vs-secondary tradeoff explanation baseline, `M8.35` as the adjacent-history drift explanation baseline, `M8.36` as the adjacent-history stability explanation baseline, and `M8.37` as the adjacent-history stability-signal baseline,
 - keep the remaining near-term budget on deeper personalization and operator rollup quality now that the continuity-explanation gap is closed on the active history surface,
 - keep remote `main` CI interpretation evidence-based: the latest fully completed `main` workflow set is green, and new pushes should be treated as pending until their workflow set finishes.
 
-## Current Mainline Snapshot (2026-04-21 Baseline Split: Foundation M8.18 / Conversation M8.36)
+## Current Mainline Snapshot (2026-04-21 Baseline Split: Foundation M8.18 / Conversation M8.37)
 
 - Foundation baseline on `main` is currently:
   - status `integrated`
@@ -37,12 +37,12 @@ Current top priority is **conversation-product hardening on top of the already-g
 - Historical increment sections below remain chronological archive entries.
 - When a historical section says `in_progress` / `no-go`, read it as past checkpoint truth rather than current repo state.
 
-## Reconciliation Notes (2026-04-21 Post-M8.36)
+## Reconciliation Notes (2026-04-21 Post-M8.37)
 
 - The active mainline question is no longer whether backend sufficiency work should continue.
   That lane is closed-by-default while its gate stays green.
-- The active mainline question is also no longer whether tradeoff, drift, and stability explanation should be added.
-  `M8.34`, `M8.35`, and `M8.36` now ship that continuity-explanation layer on the expanded history decision surface.
+- The active mainline question is also no longer whether tradeoff, drift, stability explanation, and stability signaling should be added.
+  `M8.34`, `M8.35`, `M8.36`, and `M8.37` now ship that continuity-explanation layer on the expanded history decision surface.
 - The next gap is no longer local explanation opacity.
   The remaining product question is how much deeper personalization and operator rollup value should sit on top of this now-readable current-atom history surface.
 - Remote `main` CI is currently green on the latest push.
@@ -235,6 +235,22 @@ Execution anchor:
   - diagnostics snapshot coverage for stored preview payloads,
   - EN/ZH re-render coverage for localized preview text.
 
+## Latest Mainline Increment (2026-04-21 M8.37 Active-Atom Adjacent History Stability Signals)
+
+- Extended the expanded history confidence surface with adjacent-history stability signals:
+  - stability is no longer paragraph-only; stable follow-up streaks now surface as compact badges on the same expanded history item,
+  - confidence copy now states when the current recommendation was reinforced across consecutive newer history events,
+  - the rail now distinguishes a merely fresh suggestion from one that has already held through repeated nearby history.
+- Kept the slice frontend-bounded:
+  - no backend route, API contract, persistence model, or capability schema changed,
+  - signal derivation reuses the shipped adjacent-history stability count from the same bounded recent-history window.
+- Closed the next post-`M8.36` personalization gap:
+  - operators no longer need to read the full stability paragraph to tell whether a next step has a multi-event hold,
+  - the continuity surface now exposes both stability explanation and stability-strength signaling in place.
+- Expanded regression coverage for:
+  - stability badge rendering on expanded history items with multi-event hold streaks,
+  - EN/ZH contract coverage for the new stability-signal label and confidence copy.
+
 ## Latest Mainline Increment (2026-04-21 M8.36 Active-Atom Adjacent History Stability Explanation)
 
 - Extended the expanded history decision surface with adjacent-event stability explanation:
@@ -1421,8 +1437,8 @@ Note: foundation-lane gates now exist on `main`; keep them in the normal verific
 
 1. Treat backend sufficiency as closed-by-default while the gate remains green; do not reopen heavier backend work without new measured pressure.
 2. Treat remote `main` CI as evidence-bound to the latest fully completed workflow set; only reopen CI-triage narrative when a fresh failing run exists.
-3. Make the next planning round about post-`M8.36` bounded conversation-product compounding:
-   - deepen personalization on top of the now-readable primary/secondary/drift/stability explanation surface,
+3. Make the next planning round about post-`M8.37` bounded conversation-product compounding:
+   - deepen personalization on top of the now-readable primary/secondary/drift/stability explanation surface and stability signals,
    - add broader operator rollup/reporting only when it compounds the same current-atom continuity loop,
    - keep markdown-reader improvements secondary unless they strengthen the same study loop.
 4. Treat notification-threshold governance as closed after `M7.29`; do not plan `M7.30+` unless a materially new operator requirement appears.

docs/diataxis/en/explanation/local-backend-sufficiency-and-escalation-plan.md

@@ -248,7 +248,7 @@ Recommended next product order:
    - the minimum correction / feedback closure is now integrated through explicit next-step guidance on missing managed state,
    - operator-facing visibility for conversation-memory and correction actions is now integrated through persisted diagnostics summary/index counts,
    - stronger agent action surfaces around focus mode and learning path are now integrated through the active-atom quick-action rail, toolbar routing, continuity notes/path-target diagnostics, capability-taxonomy summary cards, the recent-activity history card, localized result-detail previews on that history surface, expandable history drill-down, deterministic history follow-up affordances, continuity-aware freshness ranking for those affordances, and bounded rationale/secondary-alternative rendering,
-   - `M8.34` primary-vs-secondary tradeoff explanation, `M8.35` adjacent-history drift explanation, and `M8.36` adjacent-history stability explanation are now integrated on top of the shipped signal-enabled current-atom recent-activity card,
+   - `M8.34` primary-vs-secondary tradeoff explanation, `M8.35` adjacent-history drift explanation, `M8.36` adjacent-history stability explanation, and `M8.37` adjacent-history stability signals are now integrated on top of the shipped signal-enabled current-atom recent-activity card,
    - next only deepen personalization or broader operator rollups when they clearly compound that same readable current-atom continuity loop.
 2. markdown-reader improvement:
    - only when it clearly compounds knowledge review or focus-mode study loops.

docs/diataxis/zh/explanation/agent-conversation-focus-mode-plan.md

@@ -460,10 +460,11 @@
 `M8.34` 现在又解释主建议为什么继续排在当前 secondary alternative 前面，让展开后的 history item 不再把主建议 vs 备选决策留成隐含信息。
 `M8.35` 现在又解释相邻 history event 之间的 recommendation drift，让当前 atom history surface 不再要求操作者手工对比多张展开卡片。
 `M8.36` 现在又解释相邻 history event 之间的 recommendation stability，让当前 atom history surface 也能直接说明“新事件是否继续保持同一下一步”。
+`M8.37` 现在又在这层 explanation surface 之上补上 stability signal，让连续 hold 的 next step 可以直接从 badge 与 confidence copy 读出来，而不必读完整段 stability 文案。
 
 优先顺序：
 
-1. 在当前已落地的 primary/secondary/drift/stability explanation surface 之上继续加深 personalization，
+1. 在当前已落地的 primary/secondary/drift/stability explanation surface 与 stability signal 之上继续加深 personalization，
 2. 只有当更宽的 operator rollup 能明确复利同一 current-atom continuity loop 时才继续扩展，
 3. Markdown 阅读器提升继续保持次优先级，除非它能明确加强同一学习闭环。
 

docs/diataxis/zh/explanation/development-progress-dashboard.md

@@ -6,7 +6,7 @@
 ## 范围
 
 - 聚焦对象：本地优先学习平台（摄入、检索、学习路径、导师、记忆、治理）。
-- 时间窗口：`v1.7.0` 至 2026-04-21 foundation-integration checkpoint、2026-04-21 UTC CI 复盘周期，以及 post-`M8.36` 对齐复盘。
+- 时间窗口：`v1.7.0` 至 2026-04-21 foundation-integration checkpoint、2026-04-21 UTC CI 复盘周期，以及 post-`M8.37` 对齐复盘。
 - 证据规则：每条进展结论都需映射到：
   - 契约层（`src/learning/api.ts`、`src/learning/types.ts`），
   - 路由/运行时层（`src/server.ts`），
@@ -18,12 +18,12 @@
 当前重心是 **建立在已转绿 backend 基线之上的 conversation-product 加固**：
 
 - 让 backend sufficiency 继续作为常驻 decision gate，而不是继续占用主动执行预算，
-- 持续收口 `M8.20-M8.36` 之后暴露出的 correction / product / operator 缺口，
-- 将 `M8.23` 视作 conversation-memory / correction action 的 operator-visibility 基线，将 `M8.24` 视作当前 atom action-surface 基线，将 `M8.25` 视作当前 atom pane-state continuity 基线，将 `M8.26` 视作当前 atom study-loop summary 基线，将 `M8.27` 视作当前 atom recent-result continuity 基线，将 `M8.28` 视作当前 atom result-preview 基线，将 `M8.29` 视作当前 atom history-drill-down 基线，将 `M8.30` 视作当前 atom history follow-up 基线，将 `M8.31` 视作当前 atom history freshness-ranking 基线，将 `M8.32` 视作当前 atom history rationale/alternative 基线，将 `M8.33` 视作当前 atom history confidence-signal 基线，将 `M8.34` 视作主建议 vs secondary alternative 解释基线，将 `M8.35` 视作相邻 history drift 解释基线，并将 `M8.36` 视作相邻 history stability 解释基线，
+- 持续收口 `M8.20-M8.37` 之后暴露出的 correction / product / operator 缺口，
+- 将 `M8.23` 视作 conversation-memory / correction action 的 operator-visibility 基线，将 `M8.24` 视作当前 atom action-surface 基线，将 `M8.25` 视作当前 atom pane-state continuity 基线，将 `M8.26` 视作当前 atom study-loop summary 基线，将 `M8.27` 视作当前 atom recent-result continuity 基线，将 `M8.28` 视作当前 atom result-preview 基线，将 `M8.29` 视作当前 atom history-drill-down 基线，将 `M8.30` 视作当前 atom history follow-up 基线，将 `M8.31` 视作当前 atom history freshness-ranking 基线，将 `M8.32` 视作当前 atom history rationale/alternative 基线，将 `M8.33` 视作当前 atom history confidence-signal 基线，将 `M8.34` 视作主建议 vs secondary alternative 解释基线，将 `M8.35` 视作相邻 history drift 解释基线，将 `M8.36` 视作相邻 history stability 解释基线，并将 `M8.37` 视作相邻 history stability signal 基线，
 - 把剩余近端预算继续压在更深 personalization 与 operator rollup 质量上，而不是继续重开“本地 explanation 还不够”的缺口，
 - 远端 `main` CI 解释继续基于代码事实：最近一组已完整结束的主线工作流为绿色，新推送在整组工作流结束前只应视为 pending。
 
-## 当前主线快照（2026-04-21 基线拆分：Foundation M8.18 / Conversation M8.36）
+## 当前主线快照（2026-04-21 基线拆分：Foundation M8.18 / Conversation M8.37）
 
 - `main` 上的 foundation 基线当前为：
   - 状态 `integrated`
@@ -37,12 +37,12 @@
 - 下方“主线最新增量”章节按时间顺序保留历史记录。
 - 历史段落中出现的 `in_progress` / `no-go` 代表过去 checkpoint 结论，不代表当前仓库状态。
 
-## 对齐备注（2026-04-21 Post-M8.36）
+## 对齐备注（2026-04-21 Post-M8.37）
 
 - 当前主线问题已经不再是 backend sufficiency 是否还要继续推进。
   在 gate 继续为绿的前提下，这条 lane 默认视为闭环。
-- 当前主线问题也不再是“是否还需要补 tradeoff / drift / stability explanation”。
-  `M8.34`、`M8.35` 与 `M8.36` 已经把这层 continuity explanation 落到展开的 history 决策面上。
+- 当前主线问题也不再是“是否还需要补 tradeoff / drift / stability explanation 与 stability signaling”。
+  `M8.34`、`M8.35`、`M8.36` 与 `M8.37` 已经把这层 continuity explanation 落到展开的 history 决策面上。
 - 当前剩余缺口不再是本地 explanation 不透明。
   下一步真正的问题是：在这块已可读的 current-atom history surface 之上，还要加多少 personalization 与 operator rollup 才值得。
 - 远端 `main` CI 当前在最新推送上为绿。
@@ -235,6 +235,22 @@
   - diagnostics snapshot 中 preview payload 的可见性，
   - EN/ZH 对 preview 文本的重渲覆盖。
 
+## 主线最新增量（2026-04-21 M8.37 当前 Atom 相邻 History Stability Signals）
+
+- 已把展开的 history confidence surface 继续补上相邻事件 stability signal：
+  - stability 不再只藏在段落解释里；稳定跟随的 next step 现在会在同一张展开 history item 上直接给出 compact badge，
+  - confidence copy 现在也会明确说明：当前推荐是否已经在连续更新事件中被强化，
+  - 操作者现在可以直接区分“只是 fresh 的建议”和“已经连续保持住的建议”。
+- 继续保持 frontend-bounded：
+  - 没有新增 backend route、API contract、持久化模型或 capability schema，
+  - signal 推导完全复用已落地的相邻 history stability count 与同一段有界 recent-history 窗口。
+- 关闭了 post-`M8.36` 的下一处 personalization gap：
+  - 操作者不再需要读完整段 stability 文案，才能判断 next step 是否已经形成多事件 hold，
+  - 当前 continuity surface 现在已同时暴露 stability explanation 与 stability-strength signal。
+- 回归覆盖继续扩到：
+  - 多事件稳定 hold 的展开 history item 上渲染 stability badge，
+  - EN/ZH 对新 stability-signal label 与 confidence copy 的契约覆盖。
+
 ## 主线最新增量（2026-04-21 M8.36 当前 Atom 相邻 History Stability 解释）
 
 - 已把展开的 history 决策面继续补上相邻事件 stability explanation：
@@ -1424,8 +1440,8 @@ npm test -- src/learning/KnowledgeLearningPlatform.persistence.test.ts --runInBa
 
 1. backend sufficiency 在 gate 继续为绿时默认视为已收口；没有新的 measured pressure 前，不重开更重 backend 工作。
 2. 远端 `main` CI 继续以最近一组已完整结束的工作流为事实依据；只有出现新的失败 run，才重开 CI-triage 叙事。
-3. 下一轮 planning 应只讨论 post-`M8.36` 的有界 conversation-product 复利：
-   - 在当前已可读的 primary/secondary/drift/stability explanation surface 之上继续加深 personalization，
+3. 下一轮 planning 应只讨论 post-`M8.37` 的有界 conversation-product 复利：
+   - 在当前已可读的 primary/secondary/drift/stability explanation surface 与 stability signal 之上继续加深 personalization，
    - 只有当更宽的 operator rollup / reporting 能复利同一 current-atom continuity loop 时才继续扩展，
    - Markdown 阅读器提升继续保持次优先级，除非它能直接加强同一学习闭环。
 4. notification-threshold 治理链路继续视为在 `M7.29` 后已收口；没有新的强运维需求前，不再规划 `M7.30+`。