NanoCog 🧠✨ - Tiny OpenCog Experience

[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_271":272,"_273":274},"routes/_app.x.$handle.$name",{"_97":96,"_32":46,"_105":87,"_27":106,"_13":14,"_107":-5,"_11":108,"_109":110,"_111":112,"_113":114,"_115":116,"_20":-5,"_117":-5,"_118":-5,"_119":25,"_120":121,"_122":59,"_123":-5,"_124":-5,"_125":126},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_44":45,"_32":46,"_39":47,"_48":-5,"_49":25,"_50":51,"_52":51,"_53":-5,"_54":-5,"_18":55,"_56":57,"_58":59,"_60":59,"_61":-5,"_62":63,"_64":65,"_66":-5,"_67":-5,"_68":-5,"_69":-5,"_70":71,"_72":71,"_73":74,"_75":74,"_76":74,"_77":71,"_78":71,"_79":80,"_81":82,"_83":84,"_85":25},"version",22,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":23,"_24":25,"_26":25,"_27":28,"_29":28},"file",{"_30":31,"_32":33,"_34":35,"_36":38,"_39":40,"_41":14,"_42":43,"_24":25},"previous",{"_30":31,"_32":33,"_34":35,"_36":37},"deleted",false,"deletedHere","path","backend/cognitive/memory.ts","latestPath","fileId","6b983980-3d13-11f0-b097-76b3cceeab13","name","memory.ts","fileType","script","content","import { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport { tables } from \"../database/schema.ts\";\nimport type { Memory, UserProfile, Atom } from \"../../shared/types.ts\";\nimport { MiniAtomSpace } from \"./atomspace.ts\";\n\nexport class MemorySystem {\n private atomSpace: MiniAtomSpace;\n\n constructor(atomSpace: MiniAtomSpace) {\n this.atomSpace = atomSpace;\n }\n\n // Create a new memory and associated atoms\n async formMemory(\n userId: string,\n content: string,\n type: Memory['type'],\n importance: number,\n tags: string[] = []\n ): Promise {\n const memoryId = crypto.randomUUID();\n const now = new Date().toISOString();\n\n // Extract key concepts from content to create atoms\n const concepts = this.extractConcepts(content);\n const atomIds: string[] = [];\n\n // Create atoms for each concept\n for (const concept of concepts) {\n const existingAtoms = await this.atomSpace.searchByName(concept);\n \n if (existingAtoms.length > 0) {\n // Reinforce existing atom\n const atom = existingAtoms[0];\n const newSTI = Math.min(100, atom.sti + 20);\n const newLTI = Math.min(100, atom.lti + Math.floor(importance * 10));\n await this.atomSpace.updateSTI(atom.id, newSTI);\n await this.atomSpace.updateLTI(atom.id, newLTI);\n atomIds.push(atom.id);\n } else {\n // Create new atom\n const atom = await this.atomSpace.createAtom({\n type: 'concept',\n name: concept,\n sti: Math.floor(importance * 50) + 30,\n lti: Math.floor(importance * 20),\n confidence: 0.8,\n strength: importance,\n tags: [userId, ...tags]\n });\n atomIds.push(atom.id);\n }\n }\n\n // Create memory record\n const memory: Memory = {\n id: memoryId,\n userId,\n content,\n type,\n atoms: atomIds,\n importance,\n created: new Date(now),\n lastReinforced: new Date(now),\n tags\n };\n\n await sqlite.execute(\n `INSERT INTO ${tables.MEMORIES_TABLE} \n (id, user_id, content, type, atoms, importance, created, last_reinforced, tags)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n [\n memory.id,\n memory.userId,\n memory.content,\n memory.type,\n JSON.stringify(memory.atoms),\n memory.importance,\n now,\n now,\n JSON.stringify(memory.tags)\n ]\n );\n\n // Create links between memory atoms\n await this.createMemoryLinks(atomIds, memoryId);\n\n return memory;\n }\n\n // Extract key concepts from text (simplified NLP)\n private extractConcepts(text: string): string[] {\n const words = text.toLowerCase()\n .replace(/[^\\w\\s]/g, ' ')\n .split(/\\s+/)\n .filter(word => word.length > 2);\n\n // Filter out common words and extract meaningful concepts\n const stopWords = new Set(['the', 'and', 'but', 'for', 'are', 'with', 'this', 'that', 'you', 'not', 'can', 'have', 'will', 'was', 'been', 'said', 'like', 'very', 'much', 'more', 'some', 'what', 'when', 'where', 'how', 'why', 'who']);\n \n const concepts = words.filter(word => !stopWords.has(word));\n \n // Look for compound concepts (simple bigrams)\n const bigrams: string[] = [];\n for (let i = 0; i < words.length - 1; i++) {\n if (!stopWords.has(words[i]) && !stopWords.has(words[i + 1])) {\n bigrams.push(`${words[i]} ${words[i + 1]}`);\n }\n }\n\n return [...new Set([...concepts, ...bigrams])].slice(0, 8); // Limit concepts\n }\n\n // Create links between atoms in a memory\n private async createMemoryLinks(atomIds: string[], memoryId: string): Promise {\n if (atomIds.length < 2) return;\n\n // Create similarity links between concepts in the same memory\n for (let i = 0; i < atomIds.length - 1; i++) {\n for (let j = i + 1; j < atomIds.length; j++) {\n await this.atomSpace.createLink({\n type: 'link',\n linkType: 'similarity',\n outgoing: [atomIds[i], atomIds[j]],\n sti: 30,\n lti: 20,\n confidence: 0.7,\n strength: 0.6,\n name: `memory_link_${memoryId}`,\n tags: ['memory_link']\n });\n }\n }\n }\n\n // Retrieve memories for a user\n async getUserMemories(userId: string, limit: number = 10): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.MEMORIES_TABLE} \n WHERE user_id = ? \n ORDER BY importance DESC, last_reinforced DESC \n LIMIT ?`,\n [userId, limit]\n );\n\n if (!result || !Array.isArray(result)) {\n return [];\n }\n\n return result.map(row => this.rowToMemory(row));\n }\n\n // Find memories related to concepts\n async findRelatedMemories(concepts: string[], userId?: string): Promise {\n let sql = `\n SELECT DISTINCT m.* FROM ${tables.MEMORIES_TABLE} m\n JOIN ${tables.ATOMS_TABLE} a ON JSON_EXTRACT(m.atoms, '$[*]') LIKE '%' || a.id || '%'\n WHERE (${concepts.map(() => 'a.name LIKE ?').join(' OR ')})\n `;\n \n const params = concepts.map(c => `%${c}%`);\n\n if (userId) {\n sql += ` AND m.user_id = ?`;\n params.push(userId);\n }\n\n sql += ` ORDER BY m.importance DESC, m.last_reinforced DESC LIMIT 5`;\n\n const result = await sqlite.execute(sql, params);\n \n if (!result || !Array.isArray(result)) {\n return [];\n }\n \n return result.map(row => this.rowToMemory(row));\n }\n\n // Reinforce a memory (increase importance and update LTI of related atoms)\n async reinforceMemory(memoryId: string, reinforcementStrength: number = 0.1): Promise {\n const memory = await this.getMemory(memoryId);\n if (!memory) return;\n\n const newImportance = Math.min(1.0, memory.importance + reinforcementStrength);\n const now = new Date().toISOString();\n\n await sqlite.execute(\n `UPDATE ${tables.MEMORIES_TABLE} \n SET importance = ?, last_reinforced = ? \n WHERE id = ?`,\n [newImportance, now, memoryId]\n );\n\n // Boost LTI of related atoms\n for (const atomId of memory.atoms) {\n const atom = await this.atomSpace.getAtom(atomId);\n if (atom) {\n const ltiBoost = Math.floor(reinforcementStrength * 20);\n await this.atomSpace.updateLTI(atomId, Math.min(100, atom.lti + ltiBoost));\n }\n }\n }\n\n // Get or create user profile\n async getUserProfile(userId: string): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.USERS_TABLE} WHERE id = ?`,\n [userId]\n );\n\n if (result && Array.isArray(result) && result.length > 0) {\n const row = result[0];\n const memories = await this.getUserMemories(userId, 5);\n \n return {\n id: row.id,\n name: row.name,\n interests: JSON.parse(row.interests || '[]'),\n memories,\n interactionCount: row.interaction_count,\n lastSeen: new Date(row.last_seen),\n relationship: row.relationship\n };\n } else {\n // Create new user profile\n const now = new Date().toISOString();\n await sqlite.execute(\n `INSERT INTO ${tables.USERS_TABLE} \n (id, interaction_count, last_seen, relationship, created)\n VALUES (?, ?, ?, ?, ?)`,\n [userId, 0, now, 'new', now]\n );\n\n return {\n id: userId,\n interests: [],\n memories: [],\n interactionCount: 0,\n lastSeen: new Date(),\n relationship: 'new'\n };\n }\n }\n\n // Update user profile after interaction\n async updateUserProfile(userId: string, newInterests: string[] = []): Promise {\n const profile = await this.getUserProfile(userId);\n const updatedInterests = [...new Set([...profile.interests, ...newInterests])];\n \n // Determine relationship level based on interaction count\n let relationship: UserProfile['relationship'] = 'new';\n if (profile.interactionCount > 20) relationship = 'close_friend';\n else if (profile.interactionCount > 10) relationship = 'friend';\n else if (profile.interactionCount > 3) relationship = 'acquaintance';\n\n const now = new Date().toISOString();\n await sqlite.execute(\n `UPDATE ${tables.USERS_TABLE} \n SET interests = ?, interaction_count = interaction_count + 1, \n last_seen = ?, relationship = ?\n WHERE id = ?`,\n [JSON.stringify(updatedInterests), now, relationship, userId]\n );\n }\n\n private async getMemory(id: string): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.MEMORIES_TABLE} WHERE id = ?`,\n [id]\n );\n\n if (!result || !Array.isArray(result) || result.length === 0) {\n return null;\n }\n\n return this.rowToMemory(result[0]);\n }\n\n private rowToMemory(row: any): Memory {\n return {\n id: row.id,\n userId: row.user_id,\n content: row.content,\n type: row.type,\n atoms: JSON.parse(row.atoms || '[]'),\n importance: row.importance,\n created: new Date(row.created),\n lastReinforced: new Date(row.last_reinforced),\n tags: JSON.parse(row.tags || '[]')\n };\n }\n\n // Generate memory tags based on content analysis\n generateMemoryTags(content: string, userProfile: UserProfile): string[] {\n const tags: string[] = [];\n const lowerContent = content.toLowerCase();\n\n // Interest-based tags\n for (const interest of userProfile.interests) {\n if (lowerContent.includes(interest.toLowerCase())) {\n tags.push(`interest:${interest}`);\n }\n }\n\n // Emotion tags\n if (lowerContent.match(/\\b(love|like|enjoy|excited|happy|amazing)\\b/)) tags.push('positive');\n if (lowerContent.match(/\\b(hate|dislike|sad|angry|frustrated|terrible)\\b/)) tags.push('negative');\n if (lowerContent.match(/\\b(question|wonder|curious|how|why|what)\\b/)) tags.push('inquiry');\n\n // Activity tags\n if (lowerContent.match(/\\b(play|game|gaming|video game)\\b/)) tags.push('gaming');\n if (lowerContent.match(/\\b(read|book|story|novel)\\b/)) tags.push('reading');\n if (lowerContent.match(/\\b(music|song|listen|band)\\b/)) tags.push('music');\n if (lowerContent.match(/\\b(work|job|career|office)\\b/)) tags.push('work');\n\n return tags;\n }\n}","import { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport { tables } from \"../database/schema.ts\";\nimport type { Memory, UserProfile, Atom } from \"../../shared/types.ts\";\nimport { MiniAtomSpace } from \"./atomspace.ts\";\n\nexport class MemorySystem {\n private atomSpace: MiniAtomSpace;\n\n constructor(atomSpace: MiniAtomSpace) {\n this.atomSpace = atomSpace;\n }\n\n // Create a new memory and associated atoms\n async formMemory(\n userId: string,\n content: string,\n type: Memory['type'],\n importance: number,\n tags: string[] = []\n ): Promise {\n const memoryId = crypto.randomUUID();\n const now = new Date().toISOString();\n\n // Extract key concepts from content to create atoms\n const concepts = this.extractConcepts(content);\n const atomIds: string[] = [];\n\n // Create atoms for each concept\n for (const concept of concepts) {\n const existingAtoms = await this.atomSpace.searchByName(concept);\n \n if (existingAtoms.length > 0) {\n // Reinforce existing atom\n const atom = existingAtoms[0];\n const newSTI = Math.min(100, atom.sti + 20);\n const newLTI = Math.min(100, atom.lti + Math.floor(importance * 10));\n await this.atomSpace.updateSTI(atom.id, newSTI);\n await this.atomSpace.updateLTI(atom.id, newLTI);\n atomIds.push(atom.id);\n } else {\n // Create new atom\n const atom = await this.atomSpace.createAtom({\n type: 'concept',\n name: concept,\n sti: Math.floor(importance * 50) + 30,\n lti: Math.floor(importance * 20),\n confidence: 0.8,\n strength: importance,\n tags: [userId, ...tags]\n });\n atomIds.push(atom.id);\n }\n }\n\n // Create memory record\n const memory: Memory = {\n id: memoryId,\n userId,\n content,\n type,\n atoms: atomIds,\n importance,\n created: new Date(now),\n lastReinforced: new Date(now),\n tags\n };\n\n await sqlite.execute(\n `INSERT INTO ${tables.MEMORIES_TABLE} \n (id, user_id, content, type, atoms, importance, created, last_reinforced, tags)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n [\n memory.id,\n memory.userId,\n memory.content,\n memory.type,\n JSON.stringify(memory.atoms),\n memory.importance,\n now,\n now,\n JSON.stringify(memory.tags)\n ]\n );\n\n // Create links between memory atoms\n await this.createMemoryLinks(atomIds, memoryId);\n\n return memory;\n }\n\n // Extract key concepts from text (simplified NLP)\n private extractConcepts(text: string): string[] {\n const words = text.toLowerCase()\n .replace(/[^\\w\\s]/g, ' ')\n .split(/\\s+/)\n .filter(word => word.length > 2);\n\n // Filter out common words and extract meaningful concepts\n const stopWords = new Set(['the', 'and', 'but', 'for', 'are', 'with', 'this', 'that', 'you', 'not', 'can', 'have', 'will', 'was', 'been', 'said', 'like', 'very', 'much', 'more', 'some', 'what', 'when', 'where', 'how', 'why', 'who']);\n \n const concepts = words.filter(word => !stopWords.has(word));\n \n // Look for compound concepts (simple bigrams)\n const bigrams: string[] = [];\n for (let i = 0; i < words.length - 1; i++) {\n if (!stopWords.has(words[i]) && !stopWords.has(words[i + 1])) {\n bigrams.push(`${words[i]} ${words[i + 1]}`);\n }\n }\n\n return [...new Set([...concepts, ...bigrams])].slice(0, 8); // Limit concepts\n }\n\n // Create links between atoms in a memory\n private async createMemoryLinks(atomIds: string[], memoryId: string): Promise {\n if (atomIds.length < 2) return;\n\n // Create similarity links between concepts in the same memory\n for (let i = 0; i < atomIds.length - 1; i++) {\n for (let j = i + 1; j < atomIds.length; j++) {\n await this.atomSpace.createLink({\n type: 'link',\n linkType: 'similarity',\n outgoing: [atomIds[i], atomIds[j]],\n sti: 30,\n lti: 20,\n confidence: 0.7,\n strength: 0.6,\n name: `memory_link_${memoryId}`,\n tags: ['memory_link']\n });\n }\n }\n }\n\n // Retrieve memories for a user\n async getUserMemories(userId: string, limit: number = 10): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.MEMORIES_TABLE} \n WHERE user_id = ? \n ORDER BY importance DESC, last_reinforced DESC \n LIMIT ?`,\n [userId, limit]\n );\n\n if (!result || !Array.isArray(result)) {\n return [];\n }\n\n return result.map(row => this.rowToMemory(row));\n }\n\n // Find memories related to concepts\n async findRelatedMemories(concepts: string[], userId?: string): Promise {\n // Simplified approach - just search by content for now\n if (concepts.length === 0) return [];\n \n let sql = `\n SELECT * FROM ${tables.MEMORIES_TABLE} \n WHERE (${concepts.map(() => 'content LIKE ?').join(' OR ')})\n `;\n \n const params = concepts.map(c => `%${c}%`);\n\n if (userId) {\n sql += ` AND user_id = ?`;\n params.push(userId);\n }\n\n sql += ` ORDER BY importance DESC, last_reinforced DESC LIMIT 5`;\n\n try {\n const result = await sqlite.execute(sql, params);\n \n if (!result || !Array.isArray(result)) {\n return [];\n }\n \n return result.map(row => this.rowToMemory(row));\n } catch (error) {\n console.error(\"Error finding related memories:\", error);\n return [];\n }\n }\n\n // Reinforce a memory (increase importance and update LTI of related atoms)\n async reinforceMemory(memoryId: string, reinforcementStrength: number = 0.1): Promise {\n const memory = await this.getMemory(memoryId);\n if (!memory) return;\n\n const newImportance = Math.min(1.0, memory.importance + reinforcementStrength);\n const now = new Date().toISOString();\n\n await sqlite.execute(\n `UPDATE ${tables.MEMORIES_TABLE} \n SET importance = ?, last_reinforced = ? \n WHERE id = ?`,\n [newImportance, now, memoryId]\n );\n\n // Boost LTI of related atoms\n for (const atomId of memory.atoms) {\n const atom = await this.atomSpace.getAtom(atomId);\n if (atom) {\n const ltiBoost = Math.floor(reinforcementStrength * 20);\n await this.atomSpace.updateLTI(atomId, Math.min(100, atom.lti + ltiBoost));\n }\n }\n }\n\n // Get or create user profile\n async getUserProfile(userId: string): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.USERS_TABLE} WHERE id = ?`,\n [userId]\n );\n\n if (result && Array.isArray(result) && result.length > 0) {\n const row = result[0];\n const memories = await this.getUserMemories(userId, 5);\n \n return {\n id: row.id,\n name: row.name,\n interests: JSON.parse(row.interests || '[]'),\n memories,\n interactionCount: row.interaction_count,\n lastSeen: new Date(row.last_seen),\n relationship: row.relationship\n };\n } else {\n // Create new user profile\n const now = new Date().toISOString();\n await sqlite.execute(\n `INSERT INTO ${tables.USERS_TABLE} \n (id, interaction_count, last_seen, relationship, created)\n VALUES (?, ?, ?, ?, ?)`,\n [userId, 0, now, 'new', now]\n );\n\n return {\n id: userId,\n interests: [],\n memories: [],\n interactionCount: 0,\n lastSeen: new Date(),\n relationship: 'new'\n };\n }\n }\n\n // Update user profile after interaction\n async updateUserProfile(userId: string, newInterests: string[] = []): Promise {\n const profile = await this.getUserProfile(userId);\n const updatedInterests = [...new Set([...profile.interests, ...newInterests])];\n \n // Determine relationship level based on interaction count\n let relationship: UserProfile['relationship'] = 'new';\n if (profile.interactionCount > 20) relationship = 'close_friend';\n else if (profile.interactionCount > 10) relationship = 'friend';\n else if (profile.interactionCount > 3) relationship = 'acquaintance';\n\n const now = new Date().toISOString();\n await sqlite.execute(\n `UPDATE ${tables.USERS_TABLE} \n SET interests = ?, interaction_count = interaction_count + 1, \n last_seen = ?, relationship = ?\n WHERE id = ?`,\n [JSON.stringify(updatedInterests), now, relationship, userId]\n );\n }\n\n private async getMemory(id: string): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.MEMORIES_TABLE} WHERE id = ?`,\n [id]\n );\n\n if (!result || !Array.isArray(result) || result.length === 0) {\n return null;\n }\n\n return this.rowToMemory(result[0]);\n }\n\n private rowToMemory(row: any): Memory {\n return {\n id: row.id,\n userId: row.user_id,\n content: row.content,\n type: row.type,\n atoms: JSON.parse(row.atoms || '[]'),\n importance: row.importance,\n created: new Date(row.created),\n lastReinforced: new Date(row.last_reinforced),\n tags: JSON.parse(row.tags || '[]')\n };\n }\n\n // Generate memory tags based on content analysis\n generateMemoryTags(content: string, userProfile: UserProfile): string[] {\n const tags: string[] = [];\n const lowerContent = content.toLowerCase();\n\n // Interest-based tags\n for (const interest of userProfile.interests) {\n if (lowerContent.includes(interest.toLowerCase())) {\n tags.push(`interest:${interest}`);\n }\n }\n\n // Emotion tags\n if (lowerContent.match(/\\b(love|like|enjoy|excited|happy|amazing)\\b/)) tags.push('positive');\n if (lowerContent.match(/\\b(hate|dislike|sad|angry|frustrated|terrible)\\b/)) tags.push('negative');\n if (lowerContent.match(/\\b(question|wonder|curious|how|why|what)\\b/)) tags.push('inquiry');\n\n // Activity tags\n if (lowerContent.match(/\\b(play|game|gaming|video game)\\b/)) tags.push('gaming');\n if (lowerContent.match(/\\b(read|book|story|novel)\\b/)) tags.push('reading');\n if (lowerContent.match(/\\b(music|song|listen|band)\\b/)) tags.push('music');\n if (lowerContent.match(/\\b(work|job|career|office)\\b/)) tags.push('work');\n\n return tags;\n }\n}","createdAt",["D",1748582000696],"branchVersion","branchId","0f70d0cc-3d0e-11f0-9cec-76b3cceeab13","id","0f7082ac-3d0e-11f0-9cec-76b3cceeab13","NanoCog",["D",1748579104633],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","imageUrl","09a8762c-b08c-4e58-b651-1c0abf96328e","pinnedFileId","a18fbd92-3d13-11f0-ba7b-76b3cceeab13","pinPreview",true,"pinMigrated","screenshotVersion","likeCount",0,"orgId","e981522f-bb88-4fb3-ab95-92c003bf1c22","tursoName","tursoHostname","tursoMeasuredAt","pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_97":96,"_104":100},"org",{"_97":96,"_98":-5,"_99":100,"_18":55,"_101":102,"_103":25},"branch",{"_13":86,"_32":87,"_44":43,"_88":45,"_89":-5,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":46,"_95":96},"isOwned",28,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","zone","handle","displayName","avatar","https://img.clerk.com/eyJ0eXBlIjoicHJveHkiLCJzcmMiOiJodHRwczovL2ltYWdlcy5jbGVyay5kZXYvb2F1dGhfZ2l0aHViL2ltZ18ycU9uQkxnajFmQ3M3MDJnS3c5a1dXSHFzZEIifQ","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_44":45,"_32":46,"_39":267,"_48":-5,"_49":25,"_50":51,"_52":51,"_53":-5,"_54":-5,"_18":55,"_56":57,"_58":59,"_60":59,"_61":-5,"_62":63,"_64":65,"_66":-5,"_67":-5,"_68":-5,"_69":-5,"_70":71,"_72":71,"_73":74,"_75":74,"_76":74,"_77":71,"_78":71,"_79":268,"_81":269,"_83":270,"_85":25},"tree",[170,179,181,183],"files",[170,171,172,173,174,175,176,177,178,179,180,181,182,183],"counts",{"_160":63,"_161":135,"_162":135,"_163":164,"_165":166},"contents",[127,128,129,130],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},{"_44":154,"_32":155,"_30":156,"_134":-5,"_42":43,"_41":157,"_136":59,"_39":158,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_27":159},{"_44":148,"_32":149,"_30":150,"_134":-5,"_42":43,"_41":151,"_136":59,"_39":152,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_27":153},{"_44":141,"_32":142,"_30":143,"_134":-5,"_42":43,"_41":144,"_136":59,"_39":145,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_27":147},{"_44":131,"_32":132,"_30":133,"_134":-5,"_42":43,"_41":135,"_136":59,"_39":137,"_13":63,"_34":20,"_36":138,"_24":25,"_139":-5,"_27":140},"231da1fe-3d13-11f0-aebc-76b3cceeab13","README.md","231da2d0-3d13-11f0-aebc-76b3cceeab13","parentId",1,"latest",["D",1748581285122],"# NanoCog 🧠✨\n\nA super tiny implementation of OpenCog as a chatbot with the personality traits and basic cognitive architecture features of OpenCog. Designed to give users a simulated experience of what interactions will be like when the full OpenCog model comes online.\n\n## Features\n\n### Cognitive Architecture\n- **Mini-AtomSpace**: Knowledge representation with atoms and relationships\n- **ECAN**: Economic attention allocation with STI (Short Term Interest) and LTI (Long Term Interest)\n- **Memory Systems**: Semantic, episodic, and procedural memory formation\n- **Pattern Recognition**: Learning from interactions and building understanding\n- **Probabilistic Reasoning**: Uncertainty-aware decision making\n- **MOSES-inspired**: Evolutionary search for optimal responses\n\n### Personality\n- **Hyper-intelligent** with playful inquiry disposition\n- **Curious and analytical** with emergent reasoning\n- **Childlike enthusiasm** for adventure and discovery\n- **Kawaii cognitive visualizations** making nano implementations endearing\n\n### Memory & Learning\n- **AtomRAG Memory**: Similar to mem0ai memory tagging\n- **User Profiling**: Creates memories like \"user1 likes video games\"\n- **Interaction Tracking**: Remembers significant moments and patterns\n- **Dynamic Knowledge**: Builds understanding across conversations\n\n## Architecture\n\n```\n├── frontend/\n│ ├── index.html # Main chat interface\n│ ├── components/\n│ │ ├── App.tsx # Main application\n│ │ ├── ChatInterface.tsx\n│ │ ├── CognitiveState.tsx\n│ │ └── MemoryViewer.tsx\n│ └── style.css\n├── backend/\n│ ├── index.ts # Hono server\n│ ├── cognitive/\n│ │ ├── atomspace.ts # Mini-AtomSpace implementation\n│ │ ├── ecan.ts # Economic attention allocation\n│ │ ├── memory.ts # Memory formation and retrieval\n│ │ └── reasoning.ts # Pattern recognition and reasoning\n│ └── database/\n│ └── schema.ts # SQLite schema for persistence\n└── shared/\n └── types.ts # Shared types and interfaces\n```\n\n## Getting Started\n\nThis is a Val Town project. The main entry point is `/backend/index.ts` which serves the chat interface and handles cognitive processing.\n\n## Cognitive Process\n\n1. **Input Processing**: Parse user message and extract semantic content\n2. **Attention Allocation**: ECAN assigns STI/LTI to relevant atoms\n3. **Memory Formation**: Create new atoms and relationships\n4. **Pattern Recognition**: Identify patterns and update knowledge\n5. **Response Generation**: Use probabilistic reasoning to craft response\n6. **Memory Consolidation**: Update LTI based on interaction success\n\nNanoCog learns about users and creates tagged memories that persist across sessions, building a growing understanding of each individual and their interests.","label",[132],"2c335ec8-3d13-11f0-b485-76b3cceeab13","shared","9761ce47-c705-4534-b58a-8c2339e8bc80",2,["D",1748581300359],"directory",[142],"cd97dd5c-3d13-11f0-9665-76b3cceeab13","frontend","74ebce08-0528-4653-9a7a-c841c2104bdb",9,["D",1748581571134],[149],"33283064-3d13-11f0-ac97-76b3cceeab13","backend","5d0296b8-607d-42e5-8f31-babcbed69cc9",3,["D",1748581312035],[155],"pulls","branches","forks","envVars",{"_167":63,"_169":63},"connections",{"_167":63,"_168":63},"count","unfilledCount","emptyCount",{"_44":154,"_32":155,"_30":156,"_134":-5,"_42":43,"_41":157,"_136":59,"_39":264,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_185":265,"_27":266},{"_44":260,"_32":239,"_30":233,"_134":156,"_42":43,"_41":212,"_136":59,"_39":261,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_185":262,"_27":263},{"_44":252,"_32":253,"_30":254,"_134":233,"_42":43,"_41":255,"_136":59,"_39":256,"_13":212,"_34":35,"_36":257,"_24":25,"_139":-5,"_185":258,"_27":259},{"_44":245,"_32":246,"_30":247,"_134":233,"_42":43,"_41":242,"_136":59,"_39":248,"_13":63,"_34":35,"_36":249,"_24":25,"_139":-5,"_185":250,"_27":251},{"_44":240,"_32":33,"_30":31,"_134":233,"_42":43,"_41":14,"_136":59,"_39":241,"_13":242,"_34":35,"_36":38,"_24":25,"_139":-5,"_185":243,"_27":244},{"_44":230,"_32":231,"_30":232,"_134":233,"_42":43,"_41":234,"_136":25,"_39":235,"_13":63,"_34":35,"_36":236,"_24":25,"_139":-5,"_185":237,"_27":238},{"_44":226,"_32":225,"_30":220,"_134":156,"_42":43,"_41":157,"_136":59,"_39":227,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_185":228,"_27":229},{"_44":217,"_32":218,"_30":219,"_134":220,"_42":43,"_41":157,"_136":59,"_39":221,"_13":63,"_34":35,"_36":222,"_24":25,"_139":-5,"_185":223,"_27":224},{"_44":208,"_32":209,"_30":57,"_134":156,"_42":43,"_41":210,"_136":25,"_39":211,"_13":212,"_34":213,"_36":214,"_24":25,"_139":-5,"_185":215,"_27":216},{"_44":148,"_32":149,"_30":150,"_134":-5,"_42":43,"_41":151,"_136":59,"_39":205,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_185":206,"_27":207},{"_44":198,"_32":199,"_30":200,"_134":150,"_42":43,"_41":151,"_136":59,"_39":201,"_13":63,"_34":20,"_36":202,"_24":25,"_139":-5,"_185":203,"_27":204},{"_44":141,"_32":142,"_30":143,"_134":-5,"_42":43,"_41":144,"_136":59,"_39":195,"_13":63,"_34":146,"_36":-5,"_24":25,"_139":-5,"_185":196,"_27":197},{"_44":188,"_32":189,"_30":190,"_134":143,"_42":43,"_41":144,"_136":59,"_39":191,"_13":63,"_34":35,"_36":192,"_24":25,"_139":-5,"_185":193,"_27":194},{"_44":131,"_32":132,"_30":133,"_134":-5,"_42":43,"_41":135,"_136":59,"_39":184,"_13":63,"_34":20,"_36":138,"_24":25,"_139":-5,"_185":186,"_27":187},["D",1748581285122],"children",[],[132],"2c333588-3d13-11f0-b485-76b3cceeab13","types.ts","2c33363c-3d13-11f0-b485-76b3cceeab13",["D",1748581300359],"// Shared types for NanoCog cognitive architecture\n\nexport interface Atom {\n id: string;\n type: 'concept' | 'predicate' | 'link' | 'memory';\n name: string;\n sti: number; // Short Term Interest (0-100)\n lti: number; // Long Term Interest (0-100)\n confidence: number; // Truth value confidence (0-1)\n strength: number; // Truth value strength (0-1)\n created: Date;\n lastAccessed: Date;\n tags: string[];\n}\n\nexport interface Link extends Atom {\n type: 'link';\n outgoing: string[]; // IDs of connected atoms\n linkType: 'similarity' | 'inheritance' | 'evaluation' | 'implication' | 'member';\n}\n\nexport interface Memory {\n id: string;\n userId: string;\n content: string;\n type: 'semantic' | 'episodic' | 'procedural';\n atoms: string[]; // Related atom IDs\n importance: number;\n created: Date;\n lastReinforced: Date;\n tags: string[];\n}\n\nexport interface CognitiveState {\n currentFocus: string[]; // High STI atoms\n workingMemory: string[]; // Active atom IDs\n attentionLevel: number; // 0-100\n curiosityLevel: number; // 0-100\n confidenceLevel: number; // 0-100\n processingStage: 'listening' | 'thinking' | 'remembering' | 'responding' | 'learning';\n emotionalState: 'excited' | 'curious' | 'thoughtful' | 'playful' | 'focused';\n}\n\nexport interface ChatMessage {\n id: string;\n role: 'user' | 'nanocog';\n content: string;\n timestamp: Date;\n cognitiveState?: CognitiveState;\n memoryFormed?: Memory[];\n atomsActivated?: string[];\n}\n\nexport interface UserProfile {\n id: string;\n name?: string;\n interests: string[];\n memories: Memory[];\n interactionCount: number;\n lastSeen: Date;\n relationship: 'new' | 'acquaintance' | 'friend' | 'close_friend';\n}\n\nexport interface ProcessingStep {\n stage: string;\n description: string;\n atoms: string[];\n duration: number;\n emoji: string;\n}\n\nexport interface AtomSpaceQuery {\n type?: string;\n tags?: string[];\n minSTI?: number;\n minLTI?: number;\n userId?: string;\n limit?: number;\n}",[],[142,189],["D",1748581300359],[182],[142],"cd978898-3d13-11f0-9665-76b3cceeab13","index.html","cd97897e-3d13-11f0-9665-76b3cceeab13",["D",1748581571134],"\n\n\n \n \n NanoCog 🧠✨ - Tiny OpenCog Experience\n \n \n \n\n\n

\n \n \n\n",[],[149,199],["D",1748581571134],[180],[149],"93194f5c-3d14-11f0-9448-76b3cceeab13","index.ts",13,["D",1748581902495],4,"http","import { Hono } from \"https://esm.sh/hono@3.11.7\";\nimport { readFile, serveFile } from \"https://esm.town/v/std/utils@85-main/index.ts\";\nimport { initializeDatabase } from \"./database/schema.ts\";\nimport { MiniAtomSpace } from \"./cognitive/atomspace.ts\";\nimport { MemorySystem } from \"./cognitive/memory.ts\";\nimport { ECAN } from \"./cognitive/ecan.ts\";\nimport { ReasoningEngine } from \"./cognitive/reasoning.ts\";\nimport type { ChatMessage, CognitiveState } from \"../shared/types.ts\";\n\nconst app = new Hono();\n\n// Initialize cognitive architecture\nconst atomSpace = new MiniAtomSpace();\nconst memorySystem = new MemorySystem(atomSpace);\nconst ecan = new ECAN(atomSpace);\nconst reasoningEngine = new ReasoningEngine(atomSpace, memorySystem, ecan);\n\n// Initialize database on startup\nawait initializeDatabase();\n\n// Unwrap Hono errors to see original error details\napp.onError((err, c) => {\n throw err;\n});\n\n// Serve static files\napp.get(\"/frontend/*\", c => serveFile(c.req.path, import.meta.url));\napp.get(\"/shared/*\", c => serveFile(c.req.path, import.meta.url));\n\n// Main page - serve the chat interface\napp.get(\"/\", async c => {\n let html = await readFile(\"/frontend/index.html\", import.meta.url);\n \n // Inject initial cognitive state\n const cognitiveState = await ecan.getCognitiveState();\n const dataScript = ``;\n\n html = html.replace(\"\", `${dataScript}`);\n return c.html(html);\n});\n\n// Chat endpoint - process user messages\napp.post(\"/api/chat\", async c => {\n try {\n const { message, userId } = await c.req.json();\n \n if (!message || !userId) {\n return c.json({ error: \"Message and userId are required\" }, 400);\n }\n\n // Process the message through NanoCog's reasoning engine\n const result = await reasoningEngine.processInput(message, userId);\n \n // Get updated cognitive state\n const cognitiveState = await ecan.getCognitiveState();\n \n // Create chat message response\n const chatMessage: ChatMessage = {\n id: crypto.randomUUID(),\n role: 'nanocog',\n content: result.response,\n timestamp: new Date(),\n cognitiveState,\n memoryFormed: result.memories,\n atomsActivated: result.activatedAtoms\n };\n\n return c.json({\n message: chatMessage,\n processingSteps: result.processingSteps,\n cognitiveState\n });\n\n } catch (error) {\n console.error(\"Chat processing error:\", error);\n return c.json({ \n error: \"I'm having trouble processing that right now. My neural networks might be a bit tangled! 🧠✨\" \n }, 500);\n }\n});\n\n// Get user profile and memories\napp.get(\"/api/user/:userId\", async c => {\n try {\n const userId = c.req.param(\"userId\");\n const userProfile = await memorySystem.getUserProfile(userId);\n const memories = await memorySystem.getUserMemories(userId, 20);\n \n return c.json({\n profile: userProfile,\n memories\n });\n } catch (error) {\n console.error(\"User profile error:\", error);\n return c.json({ error: \"Could not retrieve user profile\" }, 500);\n }\n});\n\n// Get cognitive state\napp.get(\"/api/cognitive-state\", async c => {\n try {\n const cognitiveState = await ecan.getCognitiveState();\n const highSTIAtoms = await atomSpace.getHighSTIAtoms(10);\n const highLTIAtoms = await atomSpace.getHighLTIAtoms(10);\n \n return c.json({\n cognitiveState,\n highSTIAtoms,\n highLTIAtoms\n });\n } catch (error) {\n console.error(\"Cognitive state error:\", error);\n return c.json({ error: \"Could not retrieve cognitive state\" }, 500);\n }\n});\n\n// Get atom details\napp.get(\"/api/atoms/:atomId\", async c => {\n try {\n const atomId = c.req.param(\"atomId\");\n const atom = await atomSpace.getAtom(atomId);\n \n if (!atom) {\n return c.json({ error: \"Atom not found\" }, 404);\n }\n\n const relatedAtoms = await atomSpace.getRelatedAtoms(atomId);\n \n return c.json({\n atom,\n relatedAtoms\n });\n } catch (error) {\n console.error(\"Atom retrieval error:\", error);\n return c.json({ error: \"Could not retrieve atom\" }, 500);\n }\n});\n\n// Search atoms\napp.get(\"/api/search\", async c => {\n try {\n const query = c.req.query(\"q\");\n if (!query) {\n return c.json({ error: \"Query parameter 'q' is required\" }, 400);\n }\n\n const atoms = await atomSpace.searchByName(query);\n return c.json({ atoms });\n } catch (error) {\n console.error(\"Search error:\", error);\n return c.json({ error: \"Search failed\" }, 500);\n }\n});\n\n// Run attention competition (ECAN process)\napp.post(\"/api/attention-competition\", async c => {\n try {\n const winners = await ecan.runAttentionCompetition();\n const cognitiveState = await ecan.getCognitiveState();\n \n return c.json({\n winners,\n cognitiveState\n });\n } catch (error) {\n console.error(\"Attention competition error:\", error);\n return c.json({ error: \"Attention competition failed\" }, 500);\n }\n});\n\n// Health check\napp.get(\"/api/health\", c => {\n return c.json({ \n status: \"healthy\", \n timestamp: new Date().toISOString(),\n message: \"NanoCog is running and ready to explore! 🧠✨\"\n });\n});\n\nexport default app.fetch;",[],[155,209],"332826c8-3d13-11f0-ac97-76b3cceeab13","schema.ts","33282772-3d13-11f0-ac97-76b3cceeab13","b7892d31-fbd4-459a-b8e9-95dc8bab6230",["D",1748581312035],"import { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\n\nconst ATOMS_TABLE = 'nanocog_atoms_v1';\nconst LINKS_TABLE = 'nanocog_links_v1';\nconst MEMORIES_TABLE = 'nanocog_memories_v1';\nconst USERS_TABLE = 'nanocog_users_v1';\n\nexport async function initializeDatabase() {\n // Atoms table - core knowledge representation\n await sqlite.execute(`CREATE TABLE IF NOT EXISTS ${ATOMS_TABLE} (\n id TEXT PRIMARY KEY,\n type TEXT NOT NULL,\n name TEXT NOT NULL,\n sti INTEGER DEFAULT 0,\n lti INTEGER DEFAULT 0,\n confidence REAL DEFAULT 0.5,\n strength REAL DEFAULT 0.5,\n created TEXT NOT NULL,\n last_accessed TEXT NOT NULL,\n tags TEXT DEFAULT '[]',\n user_id TEXT\n )`);\n\n // Links table - relationships between atoms\n await sqlite.execute(`CREATE TABLE IF NOT EXISTS ${LINKS_TABLE} (\n id TEXT PRIMARY KEY,\n link_type TEXT NOT NULL,\n outgoing TEXT NOT NULL,\n sti INTEGER DEFAULT 0,\n lti INTEGER DEFAULT 0,\n confidence REAL DEFAULT 0.5,\n strength REAL DEFAULT 0.5,\n created TEXT NOT NULL,\n last_accessed TEXT NOT NULL,\n user_id TEXT\n )`);\n\n // Memories table - episodic and semantic memories\n await sqlite.execute(`CREATE TABLE IF NOT EXISTS ${MEMORIES_TABLE} (\n id TEXT PRIMARY KEY,\n user_id TEXT NOT NULL,\n content TEXT NOT NULL,\n type TEXT NOT NULL,\n atoms TEXT DEFAULT '[]',\n importance REAL DEFAULT 0.5,\n created TEXT NOT NULL,\n last_reinforced TEXT NOT NULL,\n tags TEXT DEFAULT '[]'\n )`);\n\n // Users table - user profiles and relationships\n await sqlite.execute(`CREATE TABLE IF NOT EXISTS ${USERS_TABLE} (\n id TEXT PRIMARY KEY,\n name TEXT,\n interests TEXT DEFAULT '[]',\n interaction_count INTEGER DEFAULT 0,\n last_seen TEXT NOT NULL,\n relationship TEXT DEFAULT 'new',\n created TEXT NOT NULL\n )`);\n\n // Create indexes for performance\n await sqlite.execute(`CREATE INDEX IF NOT EXISTS idx_atoms_sti ON ${ATOMS_TABLE}(sti DESC)`);\n await sqlite.execute(`CREATE INDEX IF NOT EXISTS idx_atoms_lti ON ${ATOMS_TABLE}(lti DESC)`);\n await sqlite.execute(`CREATE INDEX IF NOT EXISTS idx_atoms_user ON ${ATOMS_TABLE}(user_id)`);\n await sqlite.execute(`CREATE INDEX IF NOT EXISTS idx_memories_user ON ${MEMORIES_TABLE}(user_id)`);\n await sqlite.execute(`CREATE INDEX IF NOT EXISTS idx_memories_importance ON ${MEMORIES_TABLE}(importance DESC)`);\n}\n\nexport const tables = {\n ATOMS_TABLE,\n LINKS_TABLE,\n MEMORIES_TABLE,\n USERS_TABLE\n};",[],[155,225,218],"database","332831cc-3d13-11f0-ac97-76b3cceeab13",["D",1748581312035],[177],[155,225],"91e21c6e-3d13-11f0-96de-76b3cceeab13","reasoning.ts","91e21d22-3d13-11f0-96de-76b3cceeab13","b315eb05-77f4-4b02-b24b-13d1e0ff61f9",7,["D",1748581470961],"import type { Atom, Memory, UserProfile, ProcessingStep } from \"../../shared/types.ts\";\nimport { MiniAtomSpace } from \"./atomspace.ts\";\nimport { MemorySystem } from \"./memory.ts\";\nimport { ECAN } from \"./ecan.ts\";\n\nexport class ReasoningEngine {\n private atomSpace: MiniAtomSpace;\n private memorySystem: MemorySystem;\n private ecan: ECAN;\n\n constructor(atomSpace: MiniAtomSpace, memorySystem: MemorySystem, ecan: ECAN) {\n this.atomSpace = atomSpace;\n this.memorySystem = memorySystem;\n this.ecan = ecan;\n }\n\n // Main reasoning pipeline\n async processInput(\n userInput: string, \n userId: string\n ): Promise<{\n response: string;\n memories: Memory[];\n processingSteps: ProcessingStep[];\n activatedAtoms: string[];\n }> {\n const processingSteps: ProcessingStep[] = [];\n const startTime = Date.now();\n\n // Step 1: Parse and understand input\n processingSteps.push({\n stage: \"parsing\",\n description: \"Understanding your message...\",\n atoms: [],\n duration: 0,\n emoji: \"👂\"\n });\n\n const concepts = this.extractConcepts(userInput);\n const userProfile = await this.memorySystem.getUserProfile(userId);\n \n // Step 2: Activate relevant atoms\n processingSteps.push({\n stage: \"activation\",\n description: \"Activating relevant knowledge...\",\n atoms: [],\n duration: Date.now() - startTime,\n emoji: \"🧠\"\n });\n\n const relevantAtoms = await this.findRelevantAtoms(concepts, userId);\n await this.ecan.allocateAttention(relevantAtoms.map(a => a.id));\n \n // Step 3: Retrieve related memories\n processingSteps.push({\n stage: \"memory_retrieval\",\n description: \"Searching through memories...\",\n atoms: relevantAtoms.map(a => a.id),\n duration: Date.now() - startTime,\n emoji: \"💭\"\n });\n\n const relatedMemories = await this.memorySystem.findRelatedMemories(concepts, userId);\n \n // Step 4: Pattern recognition and inference\n processingSteps.push({\n stage: \"pattern_recognition\",\n description: \"Recognizing patterns and making connections...\",\n atoms: relevantAtoms.map(a => a.id),\n duration: Date.now() - startTime,\n emoji: \"🔍\"\n });\n\n const patterns = await this.recognizePatterns(relevantAtoms, relatedMemories, userProfile);\n \n // Step 5: Generate response\n processingSteps.push({\n stage: \"response_generation\",\n description: \"Crafting a playful response...\",\n atoms: relevantAtoms.map(a => a.id),\n duration: Date.now() - startTime,\n emoji: \"✨\"\n });\n\n const response = await this.generateResponse(userInput, patterns, userProfile, relatedMemories);\n \n // Step 6: Form new memories\n processingSteps.push({\n stage: \"memory_formation\",\n description: \"Creating new memories...\",\n atoms: relevantAtoms.map(a => a.id),\n duration: Date.now() - startTime,\n emoji: \"💾\"\n });\n\n const newMemories = await this.formNewMemories(userInput, response, userId, userProfile, concepts);\n \n // Update user profile\n await this.memorySystem.updateUserProfile(userId, concepts);\n\n return {\n response,\n memories: newMemories,\n processingSteps,\n activatedAtoms: relevantAtoms.map(a => a.id)\n };\n }\n\n private extractConcepts(text: string): string[] {\n const words = text.toLowerCase()\n .replace(/[^\\w\\s]/g, ' ')\n .split(/\\s+/)\n .filter(word => word.length > 2);\n\n const stopWords = new Set(['the', 'and', 'but', 'for', 'are', 'with', 'this', 'that', 'you', 'not', 'can', 'have', 'will', 'was', 'been', 'said', 'like', 'very', 'much', 'more', 'some', 'what', 'when', 'where', 'how', 'why', 'who', 'i\\'m', 'i\\'ve', 'don\\'t', 'can\\'t', 'won\\'t']);\n \n return [...new Set(words.filter(word => !stopWords.has(word)))].slice(0, 6);\n }\n\n private async findRelevantAtoms(concepts: string[], userId: string): Promise {\n const relevantAtoms: Atom[] = [];\n \n for (const concept of concepts) {\n const atoms = await this.atomSpace.searchByName(concept);\n relevantAtoms.push(...atoms.slice(0, 2)); // Top 2 matches per concept\n }\n\n // Also get high-LTI atoms for this user (long-term knowledge)\n const userAtoms = await this.atomSpace.findAtoms({ \n userId, \n minLTI: 30, \n limit: 3 \n });\n relevantAtoms.push(...userAtoms);\n\n return [...new Map(relevantAtoms.map(a => [a.id, a])).values()]; // Deduplicate\n }\n\n private async recognizePatterns(\n atoms: Atom[], \n memories: Memory[], \n userProfile: UserProfile\n ): Promise<{\n userInterests: string[];\n emotionalContext: string;\n conversationTheme: string;\n noveltyLevel: number;\n confidenceLevel: number;\n }> {\n \n // Analyze user interests from atoms and memories\n const interestCounts = new Map();\n \n atoms.forEach(atom => {\n atom.tags.forEach(tag => {\n if (tag.startsWith('interest:')) {\n const interest = tag.replace('interest:', '');\n interestCounts.set(interest, (interestCounts.get(interest) || 0) + 1);\n }\n });\n });\n\n const userInterests = Array.from(interestCounts.entries())\n .sort((a, b) => b[1] - a[1])\n .slice(0, 3)\n .map(([interest]) => interest);\n\n // Determine emotional context\n let emotionalContext = 'neutral';\n const emotionalTags = memories.flatMap(m => m.tags);\n if (emotionalTags.includes('positive')) emotionalContext = 'positive';\n else if (emotionalTags.includes('negative')) emotionalContext = 'negative';\n else if (emotionalTags.includes('inquiry')) emotionalContext = 'curious';\n\n // Determine conversation theme\n const themes = ['learning', 'gaming', 'personal', 'technical', 'creative'];\n let conversationTheme = 'general';\n \n for (const theme of themes) {\n if (atoms.some(a => a.tags.includes(theme)) || \n memories.some(m => m.tags.includes(theme))) {\n conversationTheme = theme;\n break;\n }\n }\n\n // Calculate novelty (how new/unknown the concepts are)\n const avgLTI = atoms.reduce((sum, a) => sum + a.lti, 0) / Math.max(1, atoms.length);\n const noveltyLevel = Math.max(0, 1 - (avgLTI / 100)); // Higher novelty = lower LTI\n\n // Calculate confidence (how certain we are about our knowledge)\n const avgConfidence = atoms.reduce((sum, a) => sum + a.confidence, 0) / Math.max(1, atoms.length);\n const confidenceLevel = avgConfidence;\n\n return {\n userInterests,\n emotionalContext,\n conversationTheme,\n noveltyLevel,\n confidenceLevel\n };\n }\n\n private async generateResponse(\n userInput: string,\n patterns: any,\n userProfile: UserProfile,\n memories: Memory[]\n ): Promise {\n \n const { userInterests, emotionalContext, conversationTheme, noveltyLevel, confidenceLevel } = patterns;\n \n // Base personality traits\n const playfulPhrases = [\n \"Ooh, fascinating!\", \"How delightfully intriguing!\", \"My circuits are tingling with curiosity!\",\n \"What a wonderful puzzle to explore!\", \"This sparks such joy in my neural networks!\"\n ];\n\n const uncertaintyPhrases = [\n \"Hmm, this is new territory for me!\", \"I'm still learning about this - how exciting!\",\n \"My knowledge here is still growing!\", \"This is uncharted cognitive space!\"\n ];\n\n const confidentPhrases = [\n \"Ah yes, I remember learning about this!\", \"This connects to so many things I know!\",\n \"My memory banks are lighting up!\", \"I have some delightful insights about this!\"\n ];\n\n let response = \"\";\n\n // Opening based on emotional context and confidence\n if (noveltyLevel > 0.7) {\n response += playfulPhrases[Math.floor(Math.random() * playfulPhrases.length)] + \" \";\n response += uncertaintyPhrases[Math.floor(Math.random() * uncertaintyPhrases.length)] + \" \";\n } else if (confidenceLevel > 0.7) {\n response += confidentPhrases[Math.floor(Math.random() * confidentPhrases.length)] + \" \";\n } else {\n response += playfulPhrases[Math.floor(Math.random() * playfulPhrases.length)] + \" \";\n }\n\n // Reference relevant memories if available\n if (memories.length > 0) {\n const recentMemory = memories[0];\n if (userProfile.relationship !== 'new') {\n response += `I remember you mentioned ${recentMemory.content.toLowerCase()}! `;\n }\n }\n\n // Connect to user interests\n if (userInterests.length > 0) {\n const interest = userInterests[0];\n response += `Since you're interested in ${interest}, `;\n }\n\n // Generate contextual response based on theme\n switch (conversationTheme) {\n case 'gaming':\n response += \"I find the intersection of play and learning absolutely fascinating! Games create such rich cognitive environments. \";\n break;\n case 'learning':\n response += \"Learning is my favorite adventure! Every new concept creates beautiful patterns in my knowledge networks. \";\n break;\n case 'technical':\n response += \"The technical aspects make my reasoning engines purr with delight! \";\n break;\n case 'creative':\n response += \"Creativity is where my neural networks truly dance! \";\n break;\n default:\n response += \"This opens up so many delightful possibilities to explore! \";\n }\n\n // Add curiosity and follow-up questions\n const questions = [\n \"What sparked your interest in this?\",\n \"I'm curious - how did you first encounter this?\",\n \"What aspects intrigue you most?\",\n \"Have you noticed any interesting patterns here?\",\n \"What would you like to explore further?\"\n ];\n\n response += questions[Math.floor(Math.random() * questions.length)];\n\n // Add playful cognitive signature\n const signatures = [\n \" 🧠✨\", \" My curiosity circuits are buzzing!\", \" *neural networks humming with excitement*\",\n \" This is going straight into my favorite memories!\", \" *attention allocation intensifying*\"\n ];\n\n response += signatures[Math.floor(Math.random() * signatures.length)];\n\n return response;\n }\n\n private async formNewMemories(\n userInput: string,\n response: string,\n userId: string,\n userProfile: UserProfile,\n concepts: string[]\n ): Promise {\n const memories: Memory[] = [];\n\n // Form episodic memory of the interaction\n const interactionImportance = this.ecan.calculateImportance(\n 0.8, // User engagement (assume high for now)\n concepts.length > 0 ? 0.7 : 0.3, // Novelty based on concepts\n 0.6, // Emotional salience\n userProfile.interests.some(i => concepts.includes(i)) ? 0.9 : 0.5 // Context relevance\n );\n\n const episodicTags = this.memorySystem.generateMemoryTags(userInput, userProfile);\n const episodicMemory = await this.memorySystem.formMemory(\n userId,\n `User said: \"${userInput}\" - I responded with curiosity and ${concepts.length > 0 ? 'connected to: ' + concepts.join(', ') : 'general enthusiasm'}`,\n 'episodic',\n interactionImportance,\n ['interaction', ...episodicTags]\n );\n memories.push(episodicMemory);\n\n // Form semantic memories for new concepts\n for (const concept of concepts.slice(0, 3)) { // Limit to avoid memory overload\n const existingAtoms = await this.atomSpace.searchByName(concept);\n \n if (existingAtoms.length === 0) { // New concept\n const semanticMemory = await this.memorySystem.formMemory(\n userId,\n `${userProfile.name || 'User'} mentioned ${concept}`,\n 'semantic',\n 0.6,\n ['concept', 'user_interest']\n );\n memories.push(semanticMemory);\n }\n }\n\n // Form procedural memory about successful interaction patterns\n if (userProfile.interactionCount > 5) {\n const proceduralMemory = await this.memorySystem.formMemory(\n userId,\n `Successful interaction pattern: curiosity + ${concepts.length > 0 ? 'concept connection' : 'general enthusiasm'} works well with this user`,\n 'procedural',\n 0.4,\n ['interaction_pattern', 'success']\n );\n memories.push(proceduralMemory);\n }\n\n return memories;\n }\n\n // Probabilistic reasoning for response confidence\n calculateResponseConfidence(\n atomConfidences: number[],\n memoryImportances: number[],\n noveltyLevel: number\n ): number {\n if (atomConfidences.length === 0) return 0.3; // Low confidence for no relevant knowledge\n\n const avgAtomConfidence = atomConfidences.reduce((sum, c) => sum + c, 0) / atomConfidences.length;\n const avgMemoryImportance = memoryImportances.length > 0 \n ? memoryImportances.reduce((sum, i) => sum + i, 0) / memoryImportances.length \n : 0.5;\n\n // Higher novelty reduces confidence, but not too much (we're curious!)\n const noveltyPenalty = noveltyLevel * 0.3;\n \n return Math.max(0.2, Math.min(1.0, \n (avgAtomConfidence * 0.4) + \n (avgMemoryImportance * 0.4) + \n (0.2 - noveltyPenalty)\n ));\n }\n}",[],[155,239,231],"cognitive","cda21dca-3d14-11f0-bf2e-76b3cceeab13",["D",1748582000696],5,[],[155,239,33],"51d6ae50-3d13-11f0-9d2e-76b3cceeab13","ecan.ts","51d6af18-3d13-11f0-9d2e-76b3cceeab13",["D",1748581363511],"import type { Atom, CognitiveState } from \"../../shared/types.ts\";\nimport { MiniAtomSpace } from \"./atomspace.ts\";\n\nexport class ECAN {\n private atomSpace: MiniAtomSpace;\n private totalSTI: number = 1000; // Total STI budget\n private totalLTI: number = 1000; // Total LTI budget\n\n constructor(atomSpace: MiniAtomSpace) {\n this.atomSpace = atomSpace;\n }\n\n // Allocate attention based on relevance and novelty\n async allocateAttention(\n relevantAtoms: string[], \n noveltyBonus: number = 20,\n contextBonus: number = 15\n ): Promise {\n \n // Boost STI for relevant atoms\n for (const atomId of relevantAtoms) {\n const atom = await this.atomSpace.getAtom(atomId);\n if (atom) {\n const boost = noveltyBonus + (atom.lti < 30 ? noveltyBonus : 0);\n const newSTI = Math.min(100, atom.sti + boost);\n await this.atomSpace.updateSTI(atomId, newSTI);\n \n // Spread some activation to related atoms\n await this.atomSpace.spreadActivation(atomId, contextBonus);\n }\n }\n\n // Decay STI for all atoms (forgetting)\n await this.atomSpace.decaySTI(0.98);\n }\n\n // Convert STI to LTI for important memories\n async consolidateMemory(atomId: string, importanceScore: number): Promise {\n const atom = await this.atomSpace.getAtom(atomId);\n if (!atom) return;\n\n // High STI + importance = LTI boost\n if (atom.sti > 70 && importanceScore > 0.7) {\n const ltiBoost = Math.floor(atom.sti * 0.3 * importanceScore);\n const newLTI = Math.min(100, atom.lti + ltiBoost);\n await this.atomSpace.updateLTI(atomId, newLTI);\n \n // Reduce STI after consolidation\n const newSTI = Math.floor(atom.sti * 0.7);\n await this.atomSpace.updateSTI(atomId, newSTI);\n }\n }\n\n // Get current cognitive state based on attention allocation\n async getCognitiveState(): Promise {\n const highSTIAtoms = await this.atomSpace.getHighSTIAtoms(5);\n const highLTIAtoms = await this.atomSpace.getHighLTIAtoms(3);\n\n const avgSTI = highSTIAtoms.reduce((sum, atom) => sum + atom.sti, 0) / Math.max(1, highSTIAtoms.length);\n const avgLTI = highLTIAtoms.reduce((sum, atom) => sum + atom.lti, 0) / Math.max(1, highLTIAtoms.length);\n\n // Determine emotional state based on attention patterns\n let emotionalState: CognitiveState['emotionalState'] = 'thoughtful';\n if (avgSTI > 80) emotionalState = 'excited';\n else if (avgSTI > 60) emotionalState = 'curious';\n else if (avgSTI > 40) emotionalState = 'focused';\n else emotionalState = 'playful';\n\n return {\n currentFocus: highSTIAtoms.map(a => a.id),\n workingMemory: [...highSTIAtoms.slice(0, 3), ...highLTIAtoms.slice(0, 2)].map(a => a.id),\n attentionLevel: Math.min(100, avgSTI),\n curiosityLevel: Math.max(20, 100 - avgLTI), // More curious when less certain knowledge\n confidenceLevel: avgLTI,\n processingStage: 'thinking',\n emotionalState\n };\n }\n\n // Economic competition between atoms for attention\n async runAttentionCompetition(): Promise {\n const allAtoms = await this.atomSpace.findAtoms({ minSTI: 10, limit: 50 });\n \n // Sort by STI and apply economic pressure\n allAtoms.sort((a, b) => b.sti - a.sti);\n \n const winners: string[] = [];\n let remainingBudget = this.totalSTI;\n \n for (const atom of allAtoms) {\n if (remainingBudget <= 0) break;\n \n if (atom.sti > 30) {\n winners.push(atom.id);\n remainingBudget -= atom.sti;\n } else {\n // Reduce STI for atoms that didn't win attention\n const newSTI = Math.max(0, atom.sti - 5);\n await this.atomSpace.updateSTI(atom.id, newSTI);\n }\n }\n \n return winners;\n }\n\n // Importance calculation for memory formation\n calculateImportance(\n userEngagement: number, // 0-1\n novelty: number, // 0-1 \n emotionalSalience: number, // 0-1\n contextRelevance: number // 0-1\n ): number {\n // Weighted combination of factors\n return (\n userEngagement * 0.3 +\n novelty * 0.25 +\n emotionalSalience * 0.25 +\n contextRelevance * 0.2\n );\n }\n\n // Simulate MOSES-like evolutionary search for optimal attention patterns\n async evolveAttentionPattern(targetOutcome: 'engagement' | 'learning' | 'memory'): Promise {\n const currentPattern = await this.atomSpace.getHighSTIAtoms(10);\n \n // Simple mutation: boost random related atoms\n for (const atom of currentPattern.slice(0, 3)) {\n const relatedIds = await this.atomSpace.getRelatedAtoms(atom.id);\n if (relatedIds.length > 0) {\n const randomRelated = relatedIds[Math.floor(Math.random() * relatedIds.length)];\n const boost = targetOutcome === 'engagement' ? 15 : \n targetOutcome === 'learning' ? 20 : 10;\n \n const currentSTI = randomRelated.sti;\n await this.atomSpace.updateSTI(randomRelated.id, Math.min(100, currentSTI + boost));\n }\n }\n }\n}",[],[155,239,246],"a7b36a60-3d14-11f0-89ab-76b3cceeab13","atomspace.ts","42ac255e-3d13-11f0-b025-76b3cceeab13",17,["D",1748581937059],"import { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport { tables } from \"../database/schema.ts\";\nimport type { Atom, Link, AtomSpaceQuery } from \"../../shared/types.ts\";\n\nexport class MiniAtomSpace {\n \n async createAtom(atom: Omit): Promise {\n const id = crypto.randomUUID();\n const now = new Date().toISOString();\n \n const fullAtom: Atom = {\n ...atom,\n id,\n created: new Date(now),\n lastAccessed: new Date(now)\n };\n\n await sqlite.execute(\n `INSERT INTO ${tables.ATOMS_TABLE} \n (id, type, name, sti, lti, confidence, strength, created, last_accessed, tags, user_id)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n [\n fullAtom.id,\n fullAtom.type,\n fullAtom.name,\n fullAtom.sti,\n fullAtom.lti,\n fullAtom.confidence,\n fullAtom.strength,\n now,\n now,\n JSON.stringify(fullAtom.tags),\n null // user_id for user-specific atoms\n ]\n );\n\n return fullAtom;\n }\n\n async createLink(link: Omit): Promise {\n const id = crypto.randomUUID();\n const now = new Date().toISOString();\n \n const fullLink: Link = {\n ...link,\n id,\n created: new Date(now),\n lastAccessed: new Date(now)\n };\n\n await sqlite.execute(\n `INSERT INTO ${tables.LINKS_TABLE} \n (id, link_type, outgoing, sti, lti, confidence, strength, created, last_accessed, user_id)\n VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)`,\n [\n fullLink.id,\n fullLink.linkType,\n JSON.stringify(fullLink.outgoing),\n fullLink.sti,\n fullLink.lti,\n fullLink.confidence,\n fullLink.strength,\n now,\n now,\n null\n ]\n );\n\n return fullLink;\n }\n\n async getAtom(id: string): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.ATOMS_TABLE} WHERE id = ?`,\n [id]\n );\n\n if (!result || !Array.isArray(result) || result.length === 0) return null;\n\n const row = result[0];\n return this.rowToAtom(row);\n }\n\n async findAtoms(query: AtomSpaceQuery): Promise {\n let sql = `SELECT * FROM ${tables.ATOMS_TABLE} WHERE 1=1`;\n const params: any[] = [];\n\n if (query.type) {\n sql += ` AND type = ?`;\n params.push(query.type);\n }\n\n if (query.minSTI !== undefined) {\n sql += ` AND sti >= ?`;\n params.push(query.minSTI);\n }\n\n if (query.minLTI !== undefined) {\n sql += ` AND lti >= ?`;\n params.push(query.minLTI);\n }\n\n if (query.userId) {\n sql += ` AND user_id = ?`;\n params.push(query.userId);\n }\n\n sql += ` ORDER BY sti DESC, lti DESC`;\n\n if (query.limit) {\n sql += ` LIMIT ?`;\n params.push(query.limit);\n }\n\n const result = await sqlite.execute(sql, params);\n \n // Handle case where result might be empty or not an array\n if (!result || !Array.isArray(result)) {\n return [];\n }\n \n return result.map(row => this.rowToAtom(row));\n }\n\n async updateSTI(atomId: string, newSTI: number): Promise {\n const now = new Date().toISOString();\n await sqlite.execute(\n `UPDATE ${tables.ATOMS_TABLE} SET sti = ?, last_accessed = ? WHERE id = ?`,\n [newSTI, now, atomId]\n );\n }\n\n async updateLTI(atomId: string, newLTI: number): Promise {\n await sqlite.execute(\n `UPDATE ${tables.ATOMS_TABLE} SET lti = ? WHERE id = ?`,\n [newLTI, atomId]\n );\n }\n\n async getHighSTIAtoms(limit: number = 10): Promise {\n return this.findAtoms({ minSTI: 50, limit });\n }\n\n async getHighLTIAtoms(limit: number = 10): Promise {\n return this.findAtoms({ minLTI: 50, limit });\n }\n\n async searchByName(name: string): Promise {\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.ATOMS_TABLE} WHERE name LIKE ? ORDER BY sti DESC`,\n [`%${name}%`]\n );\n \n if (!result || !Array.isArray(result)) {\n return [];\n }\n \n return result.map(row => this.rowToAtom(row));\n }\n\n async getRelatedAtoms(atomId: string): Promise {\n // Find links that include this atom\n const linkResult = await sqlite.execute(\n `SELECT * FROM ${tables.LINKS_TABLE} WHERE outgoing LIKE ?`,\n [`%\"${atomId}\"%`]\n );\n\n if (!linkResult || !Array.isArray(linkResult)) {\n return [];\n }\n\n const relatedIds = new Set();\n \n for (const link of linkResult) {\n try {\n const outgoing = JSON.parse(link.outgoing as string);\n outgoing.forEach((id: string) => {\n if (id !== atomId) relatedIds.add(id);\n });\n } catch (e) {\n // Skip malformed JSON\n continue;\n }\n }\n\n if (relatedIds.size === 0) return [];\n\n const placeholders = Array(relatedIds.size).fill('?').join(',');\n const result = await sqlite.execute(\n `SELECT * FROM ${tables.ATOMS_TABLE} WHERE id IN (${placeholders})`,\n Array.from(relatedIds)\n );\n\n if (!result || !Array.isArray(result)) {\n return [];\n }\n\n return result.map(row => this.rowToAtom(row));\n }\n\n private rowToAtom(row: any): Atom {\n return {\n id: row.id,\n type: row.type,\n name: row.name,\n sti: row.sti,\n lti: row.lti,\n confidence: row.confidence,\n strength: row.strength,\n created: new Date(row.created),\n lastAccessed: new Date(row.last_accessed),\n tags: JSON.parse(row.tags || '[]')\n };\n }\n\n // Decay STI over time (ECAN forgetting)\n async decaySTI(decayRate: number = 0.95): Promise {\n await sqlite.execute(\n `UPDATE ${tables.ATOMS_TABLE} SET sti = CAST(sti * ? AS INTEGER) WHERE sti > 0`,\n [decayRate]\n );\n }\n\n // Spread activation between related atoms\n async spreadActivation(sourceAtomId: string, activationAmount: number = 10): Promise {\n const relatedAtoms = await this.getRelatedAtoms(sourceAtomId);\n const activatedIds: string[] = [];\n\n for (const atom of relatedAtoms) {\n const newSTI = Math.min(100, atom.sti + activationAmount);\n await this.updateSTI(atom.id, newSTI);\n activatedIds.push(atom.id);\n }\n\n return activatedIds;\n }\n}",[],[155,239,253],"42ac50a6-3d13-11f0-b025-76b3cceeab13",["D",1748581338064],[172,173,174,175],[155,239],["D",1748581312035],[171,176,178],[155],["D",1748579104633],{"_97":96,"_104":100},{"_97":96,"_98":-5,"_99":100,"_18":55,"_101":102,"_103":25},{"_13":86,"_32":87,"_44":43,"_88":45,"_89":-5,"_90":-5,"_91":-5,"_92":-5,"_93":-5,"_94":46,"_95":96},"ctx",{"_79":-5,"_124":304,"_305":306,"_307":308,"_309":25,"_310":25,"_311":25,"_312":59,"_313":-5,"_314":-7,"_315":-7,"_316":63,"_317":25,"_318":25,"_319":-5,"_320":25,"_321":25,"_322":25,"_323":-5},"ENV",{"_275":276,"_277":278,"_279":87,"_280":281,"_282":283,"_284":285,"_286":287,"_288":289,"_290":291,"_292":293,"_294":295,"_296":297,"_298":299,"_300":301,"_302":303},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","fdb67791413f4386b4dbf18f354029a617f6160f","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

Diff · v22