@modusensus/dsh-mneme 0.7.31 → 0.7.32

package/src/dream/sleep.js

@@ -1,3 +1,4 @@
+import { STR, langOf } from "../lang.js";
 // System-level sleep (v0.4.0): an idle-triggered, LLM-assisted deep pass over
 // the memory store. Four independent, fail-safe phases:
 //   1. conflict resolution — high-similarity same-type pairs are either parked
@@ -15,11 +16,16 @@
 // abortable via an AbortController signal (user activity) — phases check the
 // signal between batches so a running cycle yields promptly.
 import { randomUUID, createHash } from "node:crypto";
-import { validateDecisions, applyDecisions } from "./decisions.js";
-import { findPotentialConflicts } from "./clustering.js";
-import { buildReceipt, describeStreamFailure, resolveDreamEffort, withEffortFallback } from "../dream.js";
+import { validateDecisions, applyDecisions, ACTIONS } from "./decisions.js";
+import { findPotentialConflicts, cosineSimilarity } from "./clustering.js";
+import { buildReceipt, describeStreamFailure, resolveDreamEffort, withEffortFallback, maintainIndexAfterDream } from "../dream.js";
 import { computeHeat } from "../heat.js";
 
+// Issue #126 review（Copilot）：默认档（sleepActionSet="conflict"）允许的动作 = 全局
+// ACTIONS 去掉 #126 新增的两个。从 ACTIONS 派生而非硬编码，将来再扩动作集时自动跟随。
+// 只换 prompt 挡不住模型自发输出新动作，必须由校验器按档位收口。
+const LEGACY_ACTIONS = [...ACTIONS].filter((a) => a !== "supersede" && a !== "differentiate");
+
 const SUMMARY_MAX = 120;
 // Conflict similarity threshold per strictness level (v0.4.0):
 //   gentle     only high-confidence pairs (0.92) — first-time users
@@ -27,26 +33,6 @@ const SUMMARY_MAX = 120;
 //   aggressive low-confidence pairs too (0.75) — bloated stores
 const CONFLICT_THRESHOLDS = { gentle: 0.92, normal: 0.85, aggressive: 0.75 };
 
-const CONFLICT_PROMPT = `你是记忆库冲突仲裁助手。下面是检测到的高相似度记忆对，可能内容矛盾或重复。
-对每一对输出一个 decision 对象：
-- 两条确实矛盾/重复 → { "action": "conflict", "winner": <保留的id>, "loser": <归档的id>, "reason": "理由" }
-- 两条只是主题相近、并无矛盾 → { "action": "keep", "ids": [<两个id>] }
-规则：
-- winner 应为信息更完整、更新或更可信的一条
-- 只使用提供的 id，不要编造
-- 每对必须输出一个 decision
-- 只输出 JSON 数组，不要其他文字`;
-
-const PATTERN_PROMPT = `你是记忆库模式发现助手。下面是最近的记忆条目（id、类型、标题、内容）。
-请发现跨条目的稳定模式：用户偏好的规律、反复出现的主题、可复用的工作流或项目规律。
-对每个模式输出一个 create decision：
-{ "action": "create", "type": "pattern", "title": "模式一句话标题", "content": "模式详细描述（2-4句）", "importance": 1-5, "evidence": ["支持该模式的记忆id"] }
-规则：
-- 只输出有据可依的模式，宁缺毋滥
-- evidence 必须是列表中真实存在的 id
-- 最多输出 N 个模式
-- 只输出 JSON 数组，不要其他文字`;
-
 function parseJsonArray(text) {
   if (typeof text !== "string") return undefined;
   const start = text.indexOf("[");
@@ -109,6 +95,7 @@ function makeSummary(m) {
  * each pair → winner kept / loser archived. Returns a per-run summary.
  */
 async function phaseConflicts(ctx, service, config, logger, runId, semantic = null, signal = null) {
+  const language = langOf(config);
   const embedder = semantic?.embedder;
   const vectorIndex = semantic?.vectorIndex;
   if (!embedder || !vectorIndex || typeof embedder.embed !== "function") {
@@ -150,7 +137,24 @@ async function phaseConflicts(ctx, service, config, logger, runId, semantic = nu
   const usableMemories = usable.map((i) => memories[i]);
   const usableVectors = usable.map((i) => vectors[i]);
 
+  // Issue #126：动作集档位既要决定 prompt，也要决定**校验器允许的动作集合**（见下方
+  // allowedActions），并且在候选构造阶段就要知道——full 档需要补跨类型对。因此在这里
+  // 解析，而不是等到拼 prompt 时。
+  const actionSet = config.sleepActionSet ?? "conflict";
   const pairs = findPotentialConflicts(usableMemories, usableVectors, threshold);
+  if (actionSet === "full") {
+    // Issue #126 review（Copilot）：findPotentialConflicts 只看同类型，而 full 档的
+    // prompt 明确宣传 differentiate 用于互补型——互补本来就常跨类型（同一事实的项目侧
+    // 与偏好侧）。不补这一遍，那条路径永远选不到。merge 的类型守卫不受影响：它拦在
+    // validateDecisions 里（allowCrossTypeMerge），不在候选生成。
+    for (let i = 0; i < usableMemories.length; i++) {
+      for (let j = i + 1; j < usableMemories.length; j++) {
+        if (usableMemories[i].type === usableMemories[j].type) continue;
+        const sim = cosineSimilarity(usableVectors[i], usableVectors[j]);
+        if (sim > threshold) pairs.push({ a: usableMemories[i], b: usableMemories[j], similarity: sim });
+      }
+    }
+  }
   if (pairs.length === 0) return { status: "skipped", reason: "no conflicts found" };
 
   // Dedupe: each memory participates in at most one pair, highest similarity
@@ -170,7 +174,7 @@ async function phaseConflicts(ctx, service, config, logger, runId, semantic = nu
     let frozen = 0;
     for (const p of selected) {
       try {
-        service.saveConflictPending({ run_id: runId, memory_a: p.a.id, memory_b: p.b.id, reason: `相似度 ${p.similarity.toFixed(2)}` });
+        service.saveConflictPending({ run_id: runId, memory_a: p.a.id, memory_b: p.b.id, reason: STR.similarityReason[language](p.similarity.toFixed(2)) });
         frozen++;
       } catch (error) {
         logger?.warn?.(`dsh-mneme sleep: failed to freeze conflict ${p.a.id}/${p.b.id}: ${String(error)}`);
@@ -188,10 +192,13 @@ async function phaseConflicts(ctx, service, config, logger, runId, semantic = nu
     snapshot.set(p.b.id, p.b);
   }
   const listText = selected.map((p) =>
-    `候选冲突：\nid=${p.a.id} | type=${p.a.type} | title=${p.a.title}\n${p.a.content}\n---\nid=${p.b.id} | type=${p.b.type} | title=${p.b.title}\n${p.b.content}\n（相似度 ${p.similarity.toFixed(2)}）`
+    STR.candidateConflicts[language](p)
   ).join("\n\n");
   const sleepEffort = await resolveDreamEffort(ctx, route, config.sleepReasoningEffort, logger);
   let conflictStreamFailure = "";
+  // actionSet 已在候选构造前解析（跨类型候选需要它），这里只按档位选 prompt：
+  // 默认档保持只有 conflict/keep 的窄 prompt，"full" 才开放六分支。
+  const conflictPrompt = actionSet === "full" ? STR.prompts.conflictFull[language] : STR.prompts.conflict[language];
   const runConflict = (withEffort) => {
     conflictStreamFailure = "";
     return streamText(ctx, {
@@ -201,7 +208,7 @@ async function phaseConflicts(ctx, service, config, logger, runId, semantic = nu
     maxTokens: 2048,
     ...(withEffort && sleepEffort ? { reasoningEffort: sleepEffort } : {}),
     messages: [
-      { role: "system", content: [{ type: "text", text: CONFLICT_PROMPT }] },
+      { role: "system", content: [{ type: "text", text: conflictPrompt }] },
       { role: "user", content: [{ type: "text", text: listText }] }
     ]
   }, (reason) => { conflictStreamFailure = describeStreamFailure(reason); });
@@ -218,7 +225,10 @@ async function phaseConflicts(ctx, service, config, logger, runId, semantic = nu
   // 预填作为防御性双保险——漏判读作"未裁决冲突"而非"冲突阶段整体失败"。
   const covered = new Set();
   for (const d of decisions) {
-    if (d?.action === "conflict") {
+    // Issue #126 review（Copilot）：supersede 同样用 winner/loser 定位（没有 ids），
+    // 此前只认 conflict → 它的两个目标都没进 covered → 被补成 implicit keep → 与
+    // supersede 的 claim 冲突：严格模式整单拒绝，宽容模式把合成的 keep 记成 skipped。
+    if (d?.action === "conflict" || d?.action === "supersede") {
       if (typeof d?.winner === "string") covered.add(d.winner);
       if (typeof d?.loser === "string") covered.add(d.loser);
     } else if (Array.isArray(d?.ids)) {
@@ -228,15 +238,43 @@ async function phaseConflicts(ctx, service, config, logger, runId, semantic = nu
   for (const id of snapshot.keys()) {
     if (!covered.has(id)) decisions.push({ action: "keep", ids: [id] });
   }
-  const { ok, errors } = validateDecisions(decisions, snapshot, {});
-  if (!ok) return { status: "failed", error: `invalid decisions: ${errors.join("; ")}` };
+  // Issue #126：sleep 是多对批量裁决，严格模式"一票否决"太脆（一对非法就丢掉整轮
+  // 成果）→ 与 dream 对齐透传 skipInvalid：非法的那一对只跳过自己，合法子集照常
+  // 应用，run 记 degraded。update 上限与保护期同样取 dream 的同一组配置，保证两个
+  // 模块对"什么算合法"的判据一致。
+  const { ok, errors, skipped } = validateDecisions(decisions, snapshot, {
+    maxUpdatePerRun: config.reflectionUpdateMaxPerRun,
+    minAgeHours: config.reflectionUpdateMinAgeHours,
+    skipInvalid: config.dreamSkipInvalid !== false,
+    allowCrossTypeMerge: config.allowCrossTypeMerge === true,
+    // Issue #126 review（Copilot）：默认档必须真的只放行旧动作集——只换 prompt 挡不住
+    // 模型自发输出 supersede / differentiate，那样"opt-in 零行为变化"就不成立。
+    allowedActions: actionSet === "full" ? null : LEGACY_ACTIONS
+  });
+  if (!ok) return { status: "failed", error: `invalid decisions: ${errors.join("; ")}`, skipped };
   const { applied, failures, conflicts } = applyDecisions(decisions, service, logger, snapshot, config);
+  // Issue #126 review（Copilot）：事务内 service.update 会跳过 scheduleEmbed（txDepth>0），
+  // 而 sleep 路径此前没有 dream 那样的 post-apply 索引维护——differentiate 的差异注记与
+  // supersede 的归档都反映不到缓存向量，"注记进 embedding 防再判重复"就是空承诺。
+  if (applied > 0 && semantic?.embedder && semantic?.vectorIndex) {
+    try {
+      await maintainIndexAfterDream(decisions, service, semantic);
+    } catch (error) {
+      logger?.warn?.(`dsh-mneme sleep: index maintenance failed: ${String(error)}`);
+    }
+  }
+  // Issue #126 review（Copilot）：有决策被跳过时不能报 ok——与 dream 的 degraded 契约
+  // 对齐（"合法子集已应用"是 degraded 而非 ok），否则消费方分不清整轮与残轮。
+  const degraded = skipped.length > 0;
   return {
-    status: applied > 0 ? "ok" : failures.length ? "failed" : "noop",
+    status: applied > 0 ? (degraded ? "degraded" : "ok") : failures.length ? "failed" : "noop",
     pairs: selected.length,
     applied,
     failures,
-    conflicts
+    conflicts,
+    // Issue #126：#104 同向——被跳过的逐对明细要进审计行，否则事后无法判断
+    // 是哪一对、因何被跳过（此前只进 logger.warn，splice 后就地消失）。
+    skipped
   };
 }
 
@@ -295,6 +333,7 @@ function phaseDemotion(service, config, logger, runId, signal = null) {
  * "every id claimed" invariant is trivially satisfied for pure-create lists.
  */
 async function phasePatterns(ctx, service, config, logger, runId, signal = null) {
+  const language = langOf(config);
   const route = resolveSleepRoute(ctx, config, logger);
   if (!route) return { status: "skipped", reason: "no llm route" };
   const limit = config.sleepPatternMinMemories ?? 100;
@@ -320,7 +359,7 @@ async function phasePatterns(ctx, service, config, logger, runId, signal = null)
     maxTokens: 2048,
     ...(withEffort && sleepEffort ? { reasoningEffort: sleepEffort } : {}),
     messages: [
-      { role: "system", content: [{ type: "text", text: PATTERN_PROMPT.replace("N", String(maxPatterns)) }] },
+      { role: "system", content: [{ type: "text", text: STR.prompts.pattern[language].replace("N", String(maxPatterns)) }] },
       { role: "user", content: [{ type: "text", text: listText }] }
     ]
   }, (reason) => { patternStreamFailure = describeStreamFailure(reason); });
@@ -417,7 +456,11 @@ function deriveStatus(phases) {
   const list = Object.values(phases);
   if (list.length === 0) return "noop";
   const anyError = list.some((p) => p.status === "failed" || p.status === "error");
-  const anyWork = list.some((p) => p.status === "ok");
+  // Issue #126 review（Copilot）：phase 自身可以是 degraded（合法子集已应用但有条目被
+  // 跳过），它同样算"有产出"——否则整轮会被误判成 noop，把真实变更报成空轮。
+  const anyWork = list.some((p) => p.status === "ok" || p.status === "degraded");
+  const anyDegraded = list.some((p) => p.status === "degraded");
+  if (anyDegraded) return "degraded";
   if (anyWork && anyError) return "degraded";
   if (anyError) return "failed";
   if (anyWork) return "ok";
@@ -449,6 +492,12 @@ export async function runSleep(ctx, service, config, logger, semantic = null, si
   const status = deriveStatus(phases);
   const route = resolveSleepRoute(ctx, config, logger);
   const totalApplied = Object.values(phases).reduce((n, p) => n + (Number.isInteger(p?.applied) ? p.applied : 0), 0);
+  // Issue #126（与 #104 同向）：把各 phase 的跳过明细汇总成一行审计数据，带 phase
+  // 名（同一次 run 里 conflicts 与 patterns 都可能产出 skipped）。全空时写 NULL，
+  // 不落空数组——审计行只记真实发生过的跳过。
+  const skippedDetail = Object.entries(phases).flatMap(([phase, p]) =>
+    Array.isArray(p?.skipped) ? p.skipped.map((s) => ({ phase, ...s })) : []
+  );
   const snapshotHash = createHash("sha256").update(JSON.stringify(phases)).digest("hex");
   const receipt = buildReceipt({
     runId,
@@ -473,7 +522,8 @@ export async function runSleep(ctx, service, config, logger, semantic = null, si
       summary_stored: false,
       receipt,
       policy_epoch: config.policyEpoch ?? 0,
-      run_type: "sleep"
+      run_type: "sleep",
+      skipped: skippedDetail.length ? skippedDetail : undefined
     });
   } catch (error) {
     logger?.warn?.(`dsh-mneme sleep: failed to record audit run: ${String(error)}`);