feat: Add cluster management commands for deletion, merging, and deduplication

ccw/src/core/core-memory-store.ts

@@ -522,6 +522,60 @@ ${memory.content}
     return result.changes > 0;
   }
 
+  /**
+   * Merge multiple clusters into one
+   * Keeps the first cluster and moves all members from others into it
+   * @param targetClusterId The cluster to keep
+   * @param sourceClusterIds The clusters to merge into target (will be deleted)
+   * @returns Number of members moved
+   */
+  mergeClusters(targetClusterId: string, sourceClusterIds: string[]): number {
+    const targetCluster = this.getCluster(targetClusterId);
+    if (!targetCluster) {
+      throw new Error(`Target cluster not found: ${targetClusterId}`);
+    }
+
+    let membersMoved = 0;
+    const existingMembers = new Set(
+      this.getClusterMembers(targetClusterId).map(m => m.session_id)
+    );
+
+    for (const sourceId of sourceClusterIds) {
+      if (sourceId === targetClusterId) continue;
+
+      const sourceMembers = this.getClusterMembers(sourceId);
+      const maxOrder = this.getClusterMembers(targetClusterId).length;
+
+      for (const member of sourceMembers) {
+        // Skip if already exists in target
+        if (existingMembers.has(member.session_id)) continue;
+
+        // Move member to target cluster
+        this.addClusterMember({
+          cluster_id: targetClusterId,
+          session_id: member.session_id,
+          session_type: member.session_type,
+          sequence_order: maxOrder + membersMoved + 1,
+          relevance_score: member.relevance_score
+        });
+
+        existingMembers.add(member.session_id);
+        membersMoved++;
+      }
+
+      // Delete source cluster
+      this.deleteCluster(sourceId);
+    }
+
+    // Update target cluster description
+    const finalMembers = this.getClusterMembers(targetClusterId);
+    this.updateCluster(targetClusterId, {
+      description: `Merged cluster with ${finalMembers.length} sessions`
+    });
+
+    return membersMoved;
+  }
+
   /**
    * Add member to cluster
    */
@@ -784,6 +838,219 @@ export function getCoreMemoryStore(projectPath: string): CoreMemoryStore {
   return storeCache.get(normalizedPath)!;
 }
 
+// ============================================================================
+// Cross-workspace management functions
+// ============================================================================
+
+import { readdirSync, writeFileSync, readFileSync } from 'fs';
+import { homedir } from 'os';
+
+export interface ProjectInfo {
+  id: string;
+  path: string;
+  memoriesCount: number;
+  clustersCount: number;
+  lastUpdated?: string;
+}
+
+export interface ExportedMemory {
+  version: string;
+  exportedAt: string;
+  sourceProject: string;
+  memories: CoreMemory[];
+}
+
+/**
+ * Get CCW home directory
+ */
+function getCCWHome(): string {
+  return process.env.CCW_DATA_DIR || join(homedir(), '.ccw');
+}
+
+/**
+ * List all projects with their memory counts
+ */
+export function listAllProjects(): ProjectInfo[] {
+  const projectsDir = join(getCCWHome(), 'projects');
+
+  if (!existsSync(projectsDir)) {
+    return [];
+  }
+
+  const projects: ProjectInfo[] = [];
+  const entries = readdirSync(projectsDir, { withFileTypes: true });
+
+  for (const entry of entries) {
+    if (!entry.isDirectory()) continue;
+
+    const projectId = entry.name;
+    const coreMemoryDb = join(projectsDir, projectId, 'core-memory', 'core_memory.db');
+
+    let memoriesCount = 0;
+    let clustersCount = 0;
+    let lastUpdated: string | undefined;
+
+    if (existsSync(coreMemoryDb)) {
+      try {
+        const db = new Database(coreMemoryDb, { readonly: true });
+
+        // Count memories
+        const memResult = db.prepare('SELECT COUNT(*) as count FROM memories').get() as { count: number };
+        memoriesCount = memResult?.count || 0;
+
+        // Count clusters
+        try {
+          const clusterResult = db.prepare('SELECT COUNT(*) as count FROM session_clusters').get() as { count: number };
+          clustersCount = clusterResult?.count || 0;
+        } catch {
+          // Table might not exist
+        }
+
+        // Get last update time
+        const lastMemory = db.prepare('SELECT MAX(updated_at) as last FROM memories').get() as { last: string };
+        lastUpdated = lastMemory?.last;
+
+        db.close();
+      } catch {
+        // Database might be locked or corrupted
+      }
+    }
+
+    // Convert project ID back to approximate path
+    const approximatePath = projectId
+      .replace(/^([a-z])--/, '$1:/')  // d-- -> d:/
+      .replace(/--/g, '/')
+      .replace(/-/g, ' ');
+
+    projects.push({
+      id: projectId,
+      path: approximatePath,
+      memoriesCount,
+      clustersCount,
+      lastUpdated
+    });
+  }
+
+  // Sort by last updated (most recent first)
+  return projects.sort((a, b) => {
+    if (!a.lastUpdated) return 1;
+    if (!b.lastUpdated) return -1;
+    return b.lastUpdated.localeCompare(a.lastUpdated);
+  });
+}
+
+/**
+ * Get memories from another project by ID
+ */
+export function getMemoriesFromProject(projectId: string): CoreMemory[] {
+  const projectsDir = join(getCCWHome(), 'projects');
+  const coreMemoryDb = join(projectsDir, projectId, 'core-memory', 'core_memory.db');
+
+  if (!existsSync(coreMemoryDb)) {
+    throw new Error(`Project not found: ${projectId}`);
+  }
+
+  const db = new Database(coreMemoryDb, { readonly: true });
+
+  const stmt = db.prepare('SELECT * FROM memories ORDER BY updated_at DESC');
+  const rows = stmt.all() as any[];
+
+  db.close();
+
+  return rows.map(row => ({
+    id: row.id,
+    content: row.content,
+    summary: row.summary || '',
+    raw_output: row.raw_output,
+    created_at: row.created_at,
+    updated_at: row.updated_at,
+    archived: Boolean(row.archived),
+    metadata: row.metadata
+  }));
+}
+
+/**
+ * Export memories to a JSON file
+ */
+export function exportMemories(
+  projectPath: string,
+  outputPath: string,
+  options?: { ids?: string[]; includeArchived?: boolean }
+): number {
+  const store = getCoreMemoryStore(projectPath);
+  let memories = store.getMemories({ archived: options?.includeArchived || false, limit: 10000 });
+
+  // Filter by IDs if specified
+  if (options?.ids && options.ids.length > 0) {
+    const idSet = new Set(options.ids);
+    memories = memories.filter(m => idSet.has(m.id));
+  }
+
+  const exportData: ExportedMemory = {
+    version: '1.0',
+    exportedAt: new Date().toISOString(),
+    sourceProject: projectPath,
+    memories
+  };
+
+  writeFileSync(outputPath, JSON.stringify(exportData, null, 2), 'utf-8');
+  return memories.length;
+}
+
+/**
+ * Import memories from a JSON file or another project
+ */
+export function importMemories(
+  targetProjectPath: string,
+  source: string,  // File path or project ID
+  options?: { overwrite?: boolean; prefix?: string }
+): { imported: number; skipped: number } {
+  const store = getCoreMemoryStore(targetProjectPath);
+  let memories: CoreMemory[];
+
+  // Check if source is a file or project ID
+  if (existsSync(source) && source.endsWith('.json')) {
+    // Import from file
+    const content = readFileSync(source, 'utf-8');
+    const data = JSON.parse(content) as ExportedMemory;
+    memories = data.memories;
+  } else {
+    // Import from project ID
+    memories = getMemoriesFromProject(source);
+  }
+
+  let imported = 0;
+  let skipped = 0;
+
+  for (const memory of memories) {
+    // Generate new ID with optional prefix
+    let newId = memory.id;
+    if (options?.prefix) {
+      newId = `${options.prefix}-${memory.id}`;
+    }
+
+    // Check if already exists
+    const existing = store.getMemory(newId);
+    if (existing && !options?.overwrite) {
+      skipped++;
+      continue;
+    }
+
+    // Import memory
+    store.upsertMemory({
+      id: newId,
+      content: memory.content,
+      summary: memory.summary,
+      raw_output: memory.raw_output,
+      metadata: memory.metadata
+    });
+
+    imported++;
+  }
+
+  return { imported, skipped };
+}
+
 /**
  * Close all store instances
  */

ccw/src/core/session-clustering-service.ts

@@ -488,6 +488,113 @@ export class SessionClusteringService {
     };
   }
 
+  /**
+   * Deduplicate clusters by merging similar ones
+   * Clusters with same name or >50% member overlap are merged
+   * @returns Statistics about deduplication
+   */
+  async deduplicateClusters(): Promise<{ merged: number; deleted: number; remaining: number }> {
+    const clusters = this.coreMemoryStore.listClusters('active');
+    console.log(`[Dedup] Analyzing ${clusters.length} active clusters`);
+
+    if (clusters.length < 2) {
+      return { merged: 0, deleted: 0, remaining: clusters.length };
+    }
+
+    // Group clusters by name (case-insensitive)
+    const byName = new Map<string, typeof clusters>();
+    for (const cluster of clusters) {
+      const key = cluster.name.toLowerCase().trim();
+      if (!byName.has(key)) {
+        byName.set(key, []);
+      }
+      byName.get(key)!.push(cluster);
+    }
+
+    let merged = 0;
+    let deleted = 0;
+
+    // Merge clusters with same name
+    for (const [name, group] of byName) {
+      if (group.length < 2) continue;
+
+      // Sort by created_at (oldest first) to keep the original
+      group.sort((a, b) => a.created_at.localeCompare(b.created_at));
+      const target = group[0];
+      const sources = group.slice(1).map(c => c.id);
+
+      console.log(`[Dedup] Merging ${sources.length} duplicate clusters named '${name}' into ${target.id}`);
+
+      try {
+        const membersMoved = this.coreMemoryStore.mergeClusters(target.id, sources);
+        merged += sources.length;
+        console.log(`[Dedup] Moved ${membersMoved} members, deleted ${sources.length} clusters`);
+      } catch (error) {
+        console.warn(`[Dedup] Failed to merge: ${(error as Error).message}`);
+      }
+    }
+
+    // Check for clusters with high member overlap
+    const remainingClusters = this.coreMemoryStore.listClusters('active');
+    const clusterMembers = new Map<string, Set<string>>();
+
+    for (const cluster of remainingClusters) {
+      const members = this.coreMemoryStore.getClusterMembers(cluster.id);
+      clusterMembers.set(cluster.id, new Set(members.map(m => m.session_id)));
+    }
+
+    // Find and merge overlapping clusters
+    const processed = new Set<string>();
+    for (let i = 0; i < remainingClusters.length; i++) {
+      const clusterA = remainingClusters[i];
+      if (processed.has(clusterA.id)) continue;
+
+      const membersA = clusterMembers.get(clusterA.id)!;
+      const toMerge: string[] = [];
+
+      for (let j = i + 1; j < remainingClusters.length; j++) {
+        const clusterB = remainingClusters[j];
+        if (processed.has(clusterB.id)) continue;
+
+        const membersB = clusterMembers.get(clusterB.id)!;
+        const intersection = new Set([...membersA].filter(m => membersB.has(m)));
+
+        // Calculate overlap ratio (based on smaller cluster)
+        const minSize = Math.min(membersA.size, membersB.size);
+        if (minSize > 0 && intersection.size / minSize >= 0.5) {
+          toMerge.push(clusterB.id);
+          processed.add(clusterB.id);
+        }
+      }
+
+      if (toMerge.length > 0) {
+        console.log(`[Dedup] Merging ${toMerge.length} overlapping clusters into ${clusterA.id}`);
+        try {
+          this.coreMemoryStore.mergeClusters(clusterA.id, toMerge);
+          merged += toMerge.length;
+        } catch (error) {
+          console.warn(`[Dedup] Failed to merge overlapping: ${(error as Error).message}`);
+        }
+      }
+    }
+
+    // Delete empty clusters
+    const finalClusters = this.coreMemoryStore.listClusters('active');
+    for (const cluster of finalClusters) {
+      const members = this.coreMemoryStore.getClusterMembers(cluster.id);
+      if (members.length === 0) {
+        this.coreMemoryStore.deleteCluster(cluster.id);
+        deleted++;
+        console.log(`[Dedup] Deleted empty cluster: ${cluster.id}`);
+      }
+    }
+
+    const remaining = this.coreMemoryStore.listClusters('active').length;
+    console.log(`[Dedup] Complete: ${merged} merged, ${deleted} deleted, ${remaining} remaining`);
+
+    return { merged, deleted, remaining };
+  }
+
   /**
    * Agglomerative clustering algorithm
    * Returns array of clusters (each cluster is array of sessions)