@martian-engineering/lossless-claw 0.5.3 → 0.6.1

package/src/db/config.ts

@@ -12,11 +12,14 @@ export type LcmConfig = {
   skipStatelessSessions: boolean;
   contextThreshold: number;
   freshTailCount: number;
+  newSessionRetainDepth: number;
   leafMinFanout: number;
   condensedMinFanout: number;
   condensedMinFanoutHard: number;
   incrementalMaxDepth: number;
   leafChunkTokens: number;
+  /** Maximum raw parent-history tokens imported during first-time bootstrap. */
+  bootstrapMaxTokens?: number;
   leafTargetTokens: number;
   condensedTargetTokens: number;
   maxExpandTokens: number;
@@ -39,7 +42,6 @@ export type LcmConfig = {
   expansionModel: string;
   /** Max time to wait for delegated lcm_expand_query sub-agent completion. */
   delegationTimeoutMs: number;
-  autocompactDisabled: boolean;
   /** IANA timezone for timestamps in summaries (from TZ env or system default) */
   timezone: string;
   /** When true, retroactively delete HEARTBEAT_OK turn cycles from LCM storage. */
@@ -50,6 +52,10 @@ export type LcmConfig = {
   summaryMaxOverageFactor: number;
   /** Custom instructions injected into all summarization prompts. */
   customInstructions: string;
+  /** Consecutive auth failures before the compaction circuit breaker trips (default 5). */
+  circuitBreakerThreshold: number;
+  /** Cooldown in milliseconds before the circuit breaker auto-resets (default 30 min). */
+  circuitBreakerCooldownMs: number;
 };
 
 /** Safely coerce an unknown value to a finite number, or return undefined. */
@@ -62,6 +68,21 @@ function toNumber(value: unknown): number | undefined {
   return undefined;
 }
 
+/** Safely parse a finite integer from an environment string, or return undefined.
+ *  Unlike raw parseInt(), this returns undefined for NaN so ?? fallback works. */
+function parseFiniteInt(value: string | undefined): number | undefined {
+  if (value === undefined) return undefined;
+  const parsed = parseInt(value, 10);
+  return Number.isFinite(parsed) ? parsed : undefined;
+}
+
+/** Safely parse a finite float from an environment string, or return undefined. */
+function parseFiniteNumber(value: string | undefined): number | undefined {
+  if (value === undefined) return undefined;
+  const parsed = parseFloat(value);
+  return Number.isFinite(parsed) ? parsed : undefined;
+}
+
 /** Safely coerce an unknown value to a boolean, or return undefined. */
 function toBool(value: unknown): boolean | undefined {
   if (typeof value === "boolean") return value;
@@ -108,6 +129,13 @@ export function resolveLcmConfig(
   pluginConfig?: Record<string, unknown>,
 ): LcmConfig {
   const pc = pluginConfig ?? {};
+  const resolvedLeafChunkTokens =
+    parseFiniteInt(env.LCM_LEAF_CHUNK_TOKENS)
+      ?? toNumber(pc.leafChunkTokens) ?? 20000;
+  const resolvedBootstrapMaxTokens =
+    parseFiniteInt(env.LCM_BOOTSTRAP_MAX_TOKENS)
+      ?? toNumber(pc.bootstrapMaxTokens)
+      ?? Math.max(6000, Math.floor(resolvedLeafChunkTokens * 0.3));
   const envDelegationTimeoutMs =
     env.LCM_DELEGATION_TIMEOUT_MS !== undefined
       ? toNumber(env.LCM_DELEGATION_TIMEOUT_MS)
@@ -142,37 +170,39 @@ export function resolveLcmConfig(
         ? env.LCM_SKIP_STATELESS_SESSIONS === "true"
         : toBool(pc.skipStatelessSessions) ?? true,
     contextThreshold:
-      (env.LCM_CONTEXT_THRESHOLD !== undefined ? parseFloat(env.LCM_CONTEXT_THRESHOLD) : undefined)
+      parseFiniteNumber(env.LCM_CONTEXT_THRESHOLD)
         ?? toNumber(pc.contextThreshold) ?? 0.75,
     freshTailCount:
-      (env.LCM_FRESH_TAIL_COUNT !== undefined ? parseInt(env.LCM_FRESH_TAIL_COUNT, 10) : undefined)
+      parseFiniteInt(env.LCM_FRESH_TAIL_COUNT)
         ?? toNumber(pc.freshTailCount) ?? 64,
+    newSessionRetainDepth:
+      parseFiniteInt(env.LCM_NEW_SESSION_RETAIN_DEPTH)
+        ?? toNumber(pc.newSessionRetainDepth) ?? 2,
     leafMinFanout:
-      (env.LCM_LEAF_MIN_FANOUT !== undefined ? parseInt(env.LCM_LEAF_MIN_FANOUT, 10) : undefined)
+      parseFiniteInt(env.LCM_LEAF_MIN_FANOUT)
         ?? toNumber(pc.leafMinFanout) ?? 8,
     condensedMinFanout:
-      (env.LCM_CONDENSED_MIN_FANOUT !== undefined ? parseInt(env.LCM_CONDENSED_MIN_FANOUT, 10) : undefined)
+      parseFiniteInt(env.LCM_CONDENSED_MIN_FANOUT)
         ?? toNumber(pc.condensedMinFanout) ?? 4,
     condensedMinFanoutHard:
-      (env.LCM_CONDENSED_MIN_FANOUT_HARD !== undefined ? parseInt(env.LCM_CONDENSED_MIN_FANOUT_HARD, 10) : undefined)
+      parseFiniteInt(env.LCM_CONDENSED_MIN_FANOUT_HARD)
         ?? toNumber(pc.condensedMinFanoutHard) ?? 2,
     incrementalMaxDepth:
-      (env.LCM_INCREMENTAL_MAX_DEPTH !== undefined ? parseInt(env.LCM_INCREMENTAL_MAX_DEPTH, 10) : undefined)
+      parseFiniteInt(env.LCM_INCREMENTAL_MAX_DEPTH)
         ?? toNumber(pc.incrementalMaxDepth) ?? 1,
-    leafChunkTokens:
-      (env.LCM_LEAF_CHUNK_TOKENS !== undefined ? parseInt(env.LCM_LEAF_CHUNK_TOKENS, 10) : undefined)
-        ?? toNumber(pc.leafChunkTokens) ?? 20000,
+    leafChunkTokens: resolvedLeafChunkTokens,
+    bootstrapMaxTokens: resolvedBootstrapMaxTokens,
     leafTargetTokens:
-      (env.LCM_LEAF_TARGET_TOKENS !== undefined ? parseInt(env.LCM_LEAF_TARGET_TOKENS, 10) : undefined)
+      parseFiniteInt(env.LCM_LEAF_TARGET_TOKENS)
         ?? toNumber(pc.leafTargetTokens) ?? 2400,
     condensedTargetTokens:
-      (env.LCM_CONDENSED_TARGET_TOKENS !== undefined ? parseInt(env.LCM_CONDENSED_TARGET_TOKENS, 10) : undefined)
+      parseFiniteInt(env.LCM_CONDENSED_TARGET_TOKENS)
         ?? toNumber(pc.condensedTargetTokens) ?? 2000,
     maxExpandTokens:
-      (env.LCM_MAX_EXPAND_TOKENS !== undefined ? parseInt(env.LCM_MAX_EXPAND_TOKENS, 10) : undefined)
+      parseFiniteInt(env.LCM_MAX_EXPAND_TOKENS)
         ?? toNumber(pc.maxExpandTokens) ?? 4000,
     largeFileTokenThreshold:
-      (env.LCM_LARGE_FILE_TOKEN_THRESHOLD !== undefined ? parseInt(env.LCM_LARGE_FILE_TOKEN_THRESHOLD, 10) : undefined)
+      parseFiniteInt(env.LCM_LARGE_FILE_TOKEN_THRESHOLD)
         ?? toNumber(pc.largeFileThresholdTokens)
         ?? toNumber(pc.largeFileTokenThreshold)
         ?? 25000,
@@ -189,22 +219,24 @@ export function resolveLcmConfig(
     expansionModel:
       env.LCM_EXPANSION_MODEL?.trim() ?? toStr(pc.expansionModel) ?? "",
     delegationTimeoutMs: envDelegationTimeoutMs ?? toNumber(pc.delegationTimeoutMs) ?? 120000,
-    autocompactDisabled:
-      env.LCM_AUTOCOMPACT_DISABLED !== undefined
-        ? env.LCM_AUTOCOMPACT_DISABLED === "true"
-        : toBool(pc.autocompactDisabled) ?? false,
     timezone: env.TZ ?? toStr(pc.timezone) ?? Intl.DateTimeFormat().resolvedOptions().timeZone,
     pruneHeartbeatOk:
       env.LCM_PRUNE_HEARTBEAT_OK !== undefined
         ? env.LCM_PRUNE_HEARTBEAT_OK === "true"
         : toBool(pc.pruneHeartbeatOk) ?? false,
     maxAssemblyTokenBudget:
-      (env.LCM_MAX_ASSEMBLY_TOKEN_BUDGET !== undefined ? parseInt(env.LCM_MAX_ASSEMBLY_TOKEN_BUDGET, 10) : undefined)
+      parseFiniteInt(env.LCM_MAX_ASSEMBLY_TOKEN_BUDGET)
         ?? toNumber(pc.maxAssemblyTokenBudget) ?? undefined,
     summaryMaxOverageFactor:
-      (env.LCM_SUMMARY_MAX_OVERAGE_FACTOR !== undefined ? parseFloat(env.LCM_SUMMARY_MAX_OVERAGE_FACTOR) : undefined)
+      parseFiniteNumber(env.LCM_SUMMARY_MAX_OVERAGE_FACTOR)
         ?? toNumber(pc.summaryMaxOverageFactor) ?? 3,
     customInstructions:
       env.LCM_CUSTOM_INSTRUCTIONS?.trim() ?? toStr(pc.customInstructions) ?? "",
+    circuitBreakerThreshold:
+      parseFiniteInt(env.LCM_CIRCUIT_BREAKER_THRESHOLD)
+        ?? toNumber(pc.circuitBreakerThreshold) ?? 5,
+    circuitBreakerCooldownMs:
+      parseFiniteInt(env.LCM_CIRCUIT_BREAKER_COOLDOWN_MS)
+        ?? toNumber(pc.circuitBreakerCooldownMs) ?? 1_800_000,
   };
 }

package/src/db/migration.ts

@@ -1,5 +1,6 @@
 import type { DatabaseSync } from "node:sqlite";
 import { getLcmDbFeatures } from "./features.js";
+import { parseUtcTimestampOrNull } from "../store/parse-utc-timestamp.js";
 
 type SummaryColumnInfo = {
   name?: string;
@@ -62,18 +63,7 @@ function ensureSummaryMetadataColumns(db: DatabaseSync): void {
 }
 
 function parseTimestamp(value: string | null | undefined): Date | null {
-  if (typeof value !== "string" || !value.trim()) {
-    return null;
-  }
-
-  const direct = new Date(value);
-  if (!Number.isNaN(direct.getTime())) {
-    return direct;
-  }
-
-  const normalized = value.includes("T") ? value : `${value.replace(" ", "T")}Z`;
-  const parsed = new Date(normalized);
-  return Number.isNaN(parsed.getTime()) ? null : parsed;
+  return parseUtcTimestampOrNull(value);
 }
 
 function isoStringOrNull(value: Date | null): string | null {
@@ -434,6 +424,8 @@ export function runLcmMigrations(
       conversation_id INTEGER PRIMARY KEY AUTOINCREMENT,
       session_id TEXT NOT NULL,
       session_key TEXT,
+      active INTEGER NOT NULL DEFAULT 1,
+      archived_at TEXT,
       title TEXT,
       bootstrapped_at TEXT,
       created_at TEXT NOT NULL DEFAULT (datetime('now')),
@@ -579,7 +571,27 @@ export function runLcmMigrations(
     db.exec(`ALTER TABLE conversations ADD COLUMN session_key TEXT`);
   }
 
-  db.exec(`CREATE UNIQUE INDEX IF NOT EXISTS conversations_session_key_idx ON conversations (session_key)`);
+  const hasActive = conversationColumns.some((col) => col.name === "active");
+  if (!hasActive) {
+    db.exec(`ALTER TABLE conversations ADD COLUMN active INTEGER NOT NULL DEFAULT 1`);
+  }
+
+  const hasArchivedAt = conversationColumns.some((col) => col.name === "archived_at");
+  if (!hasArchivedAt) {
+    db.exec(`ALTER TABLE conversations ADD COLUMN archived_at TEXT`);
+  }
+
+  db.exec(`UPDATE conversations SET active = 1 WHERE active IS NULL`);
+  db.exec(`
+    CREATE UNIQUE INDEX IF NOT EXISTS conversations_active_session_key_idx
+    ON conversations (session_key)
+    WHERE session_key IS NOT NULL AND active = 1
+  `);
+  db.exec(`
+    CREATE INDEX IF NOT EXISTS conversations_session_key_active_created_idx
+    ON conversations (session_key, active, created_at)
+  `);
+  db.exec(`DROP INDEX IF EXISTS conversations_session_key_idx`);
   ensureSummaryDepthColumn(db);
   ensureSummaryMetadataColumns(db);
   ensureSummaryModelColumn(db);
@@ -649,4 +661,29 @@ export function runLcmMigrations(
       SELECT summary_id, content FROM summaries;
     `);
   }
+
+  // ── CJK trigram FTS table ────────────────────────────────────────────────
+  // FTS5 unicode61 (porter) tokenizer cannot segment CJK ideographs, so CJK
+  // queries currently fall back to a LIKE path with AND logic.  When the user's
+  // phrasing doesn't match the summary verbatim (e.g. "端到端测试结果" vs
+  // "端到端测试"), ALL terms must match and the query returns 0 candidates.
+  //
+  // A trigram-tokenized table indexes every 3-character substring, enabling
+  // native CJK substring matching via FTS5 MATCH with OR semantics.
+  const cjkTableExists = db
+    .prepare(
+      "SELECT 1 FROM sqlite_master WHERE type='table' AND name='summaries_fts_cjk'",
+    )
+    .get();
+  if (!cjkTableExists) {
+    db.exec(`
+      CREATE VIRTUAL TABLE summaries_fts_cjk USING fts5(
+        summary_id UNINDEXED,
+        content,
+        tokenize='trigram'
+      );
+      INSERT INTO summaries_fts_cjk(summary_id, content)
+      SELECT summary_id, content FROM summaries;
+    `);
+  }
 }