n8n-nodes-tembory 1.0.6 → 1.0.7

package/dist/nodes/Mem0/Mem0Memory.node.js

@@ -976,6 +976,16 @@ const dedupeToolHistory = (items) => {
         const key = toolEventKey(item);
         if (seen.has(key))
             continue;
+        const atMs = new Date(item.at || 0).getTime();
+        const nearDuplicate = out.some((existing) => {
+            const existingMs = new Date(existing.at || 0).getTime();
+            return Math.abs(atMs - existingMs) < 1000
+                && String(item.name || '') === String(existing.name || '')
+                && canonicalToolInput(item.input) === canonicalToolInput(existing.input)
+                && String(item.result || '') === String(existing.result || '');
+        });
+        if (nearDuplicate)
+            continue;
         seen.add(key);
         out.push(item);
     }
@@ -1211,7 +1221,7 @@ const deriveMemoryCompression = ({ recentMessages = [], toolHistory = [], profil
     const lastMessages = pruneByLimit(recentMessages || [], maxItems).map((msg) => `${msg.role || 'message'}: ${truncate(msg.content || '', 160)}`);
     const lastTools = pruneByLimit(toolHistory || [], maxItems).map((tool) => `${tool.name || 'tool'}:${tool.ok === false ? 'failed' : 'ok'}${tool.at ? `@${tool.at}` : ''}`);
     const profile = renderProfileFacts(profileFacts);
-    const activeMemories = (vectorMemories || []).map(enrichContextMemory).filter((item) => item.status === 'active').slice(0, maxItems);
+    const activeMemories = (vectorMemories || []).map((memory) => ({ ...enrichContextMemory(memory), compact_memory: contextMemoryText(memory, 220) })).filter((item) => item.status === 'active').slice(0, maxItems);
     return {
         turn_summary: lastMessages.length ? lastMessages : [],
         session_summary: {
@@ -1223,7 +1233,7 @@ const deriveMemoryCompression = ({ recentMessages = [], toolHistory = [], profil
         workflow_summary: {
             active_memory_count: activeMemories.length,
             top_active_memories: activeMemories.map((item) => ({
-                memory: truncate(item.memory, 220),
+                memory: item.compact_memory,
                 confidence: item.confidence,
                 source: item.source,
             })),
@@ -2357,7 +2367,7 @@ class Mem0Memory {
             const rawScore = scoreOf(m);
             const enriched = enrichContextMemory(m);
             return {
-                memory: memoryText(m),
+                memory: contextMemoryText(m, Number(adv.vectorMemoryMaxChars || 700)),
                 score: adv.includeScores === false ? undefined : (rawScore !== undefined ? rawScore : scoreMeta.hybridScore ?? scoreMeta.recencyScore ?? 'unavailable'),
                 semantic_score: adv.includeScores === false ? undefined : scoreMeta.semanticScore,
                 recency_score: adv.includeScores === false ? undefined : scoreMeta.recencyScore,
@@ -2372,7 +2382,7 @@ class Mem0Memory {
                 raw: adv.includeDiagnostics ? m : undefined,
             };
         });
-        const summary = vectorMemories.slice(0, Number(adv.summaryMaxFacts || 4)).map(memoryText).filter(Boolean);
+        const summary = vectorMemories.slice(0, Number(adv.summaryMaxFacts || 4)).map((memory) => contextMemoryText(memory, 360)).filter(Boolean);
         const contextText = payload.map((message) => String(message.content || '')).join('\n\n');
         const audit = {
             kind: 'elefai.brain.context.v1',

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "n8n-nodes-tembory",
-  "version": "1.0.6",
+  "version": "1.0.7",
   "description": "Tembory node for n8n AI Agents with profile, tools, timeline, graph and semantic memory",
   "license": "MIT",
   "homepage": "https://tembory.com",