gitnexus 1.1.0 → 1.1.2

package/dist/core/ingestion/process-processor.js

@@ -15,7 +15,7 @@ const DEFAULT_CONFIG = {
     maxTraceDepth: 10,
     maxBranching: 4,
     maxProcesses: 75,
-    minSteps: 2,
+    minSteps: 3, // 3+ steps = genuine multi-hop flow (2-step is just "A calls B")
 };
 // ============================================================================
 // MAIN PROCESSOR
@@ -51,10 +51,13 @@ export const processProcesses = async (knowledgeGraph, memberships, onProgress,
         }
     }
     onProgress?.(`Found ${allTraces.length} traces, deduplicating...`, 60);
-    // Step 3: Deduplicate similar traces
+    // Step 3: Deduplicate similar traces (subset removal)
     const uniqueTraces = deduplicateTraces(allTraces);
+    // Step 3b: Deduplicate by entry+terminal pair (keep longest path per pair)
+    const endpointDeduped = deduplicateByEndpoints(uniqueTraces);
+    onProgress?.(`Deduped ${uniqueTraces.length} → ${endpointDeduped.length} unique endpoint pairs`, 70);
     // Step 4: Limit to max processes (prioritize longer traces)
-    const limitedTraces = uniqueTraces
+    const limitedTraces = endpointDeduped
         .sort((a, b) => b.length - a.length)
         .slice(0, cfg.maxProcesses);
     onProgress?.(`Creating ${limitedTraces.length} process nodes...`, 80);
@@ -266,6 +269,27 @@ const deduplicateTraces = (traces) => {
     return unique;
 };
 // ============================================================================
+// HELPER: Deduplicate by entry+terminal endpoints
+// ============================================================================
+/**
+ * Keep only the longest trace per unique entry→terminal pair.
+ * Multiple paths between the same two endpoints are redundant for agents.
+ */
+const deduplicateByEndpoints = (traces) => {
+    if (traces.length === 0)
+        return [];
+    const byEndpoints = new Map();
+    // Sort longest first so the first seen per key is the longest
+    const sorted = [...traces].sort((a, b) => b.length - a.length);
+    for (const trace of sorted) {
+        const key = `${trace[0]}::${trace[trace.length - 1]}`;
+        if (!byEndpoints.has(key)) {
+            byEndpoints.set(key, trace);
+        }
+    }
+    return Array.from(byEndpoints.values());
+};
+// ============================================================================
 // HELPER: String utilities
 // ============================================================================
 const capitalize = (s) => {

package/dist/core/search/bm25-index.js

@@ -11,11 +11,11 @@ import { queryFTS } from '../kuzu/kuzu-adapter.js';
  */
 async function queryFTSViaExecutor(executor, tableName, indexName, query, limit) {
     const escapedQuery = query.replace(/'/g, "''");
-    const cypher = `
-    CALL QUERY_FTS_INDEX('${tableName}', '${indexName}', '${escapedQuery}', conjunctive := false)
-    RETURN node, score
-    ORDER BY score DESC
-    LIMIT ${limit}
+    const cypher = `
+    CALL QUERY_FTS_INDEX('${tableName}', '${indexName}', '${escapedQuery}', conjunctive := false)
+    RETURN node, score
+    ORDER BY score DESC
+    LIMIT ${limit}
   `;
     try {
         const rows = await executor(cypher);

package/dist/mcp/local/local-backend.d.ts

@@ -11,19 +11,9 @@ export interface CodebaseContext {
     stats: {
         fileCount: number;
         functionCount: number;
-        classCount: number;
-        interfaceCount: number;
-        methodCount: number;
         communityCount: number;
         processCount: number;
     };
-    hotspots: Array<{
-        name: string;
-        type: string;
-        filePath: string;
-        connections: number;
-    }>;
-    folderTree: string;
 }
 interface RepoHandle {
     id: string;
@@ -57,24 +47,10 @@ export declare class LocalBackend {
      */
     resolveRepo(repoParam?: string): RepoHandle;
     private ensureInitialized;
-    get isReady(): boolean;
     /**
      * Get context for a specific repo (or the single repo if only one).
      */
     getContext(repoId?: string): CodebaseContext | null;
-    /**
-     * Backwards-compatible getter — returns context of single repo or null.
-     */
-    get context(): CodebaseContext | null;
-    getRepoPath(repoId?: string): string | null;
-    get repoPath(): string | null;
-    getMeta(repoId?: string): {
-        lastCommit: string;
-        indexedAt: string;
-        stats?: any;
-    } | null;
-    get meta(): any;
-    get storagePath(): string | null;
     /**
      * List all registered repos with their metadata.
      */
@@ -96,6 +72,12 @@ export declare class LocalBackend {
      */
     private semanticSearch;
     private cypher;
+    /**
+     * Aggregate same-named clusters: group by heuristicLabel, sum symbols,
+     * weighted-average cohesion, filter out tiny clusters (<5 symbols).
+     * Raw communities stay intact in KuzuDB for Cypher queries.
+     */
+    private aggregateClusters;
     private overview;
     private explore;
     private impact;

package/dist/mcp/local/local-backend.js

@@ -57,14 +57,9 @@ export class LocalBackend {
                 stats: {
                     fileCount: s.files || 0,
                     functionCount: s.nodes || 0,
-                    classCount: 0,
-                    interfaceCount: 0,
-                    methodCount: 0,
                     communityCount: s.communities || 0,
                     processCount: s.processes || 0,
                 },
-                hotspots: [],
-                folderTree: '',
             });
         }
         return this.repos.size > 0;
@@ -137,9 +132,6 @@ export class LocalBackend {
         this.initializedRepos.add(repoId);
     }
     // ─── Public Getters ──────────────────────────────────────────────
-    get isReady() {
-        return this.repos.size > 0;
-    }
     /**
      * Get context for a specific repo (or the single repo if only one).
      */
@@ -152,38 +144,6 @@ export class LocalBackend {
         }
         return null;
     }
-    /**
-     * Backwards-compatible getter — returns context of single repo or null.
-     */
-    get context() {
-        return this.getContext();
-    }
-    getRepoPath(repoId) {
-        if (repoId)
-            return this.repos.get(repoId)?.repoPath ?? null;
-        if (this.repos.size === 1)
-            return this.repos.values().next().value?.repoPath ?? null;
-        return null;
-    }
-    get repoPath() {
-        return this.getRepoPath();
-    }
-    getMeta(repoId) {
-        const handle = repoId
-            ? this.repos.get(repoId)
-            : this.repos.size === 1 ? this.repos.values().next().value : null;
-        if (!handle)
-            return null;
-        return { lastCommit: handle.lastCommit, indexedAt: handle.indexedAt, stats: handle.stats };
-    }
-    get meta() {
-        return this.getMeta();
-    }
-    get storagePath() {
-        if (this.repos.size === 1)
-            return this.repos.values().next().value?.storagePath ?? null;
-        return null;
-    }
     /**
      * List all registered repos with their metadata.
      */
@@ -274,10 +234,10 @@ export class LocalBackend {
             // Add cluster membership context for each result with a nodeId
             if (result.nodeId) {
                 try {
-                    const clusterQuery = `
-            MATCH (n {id: '${result.nodeId.replace(/'/g, "''")}'})-[:CodeRelation {type: 'MEMBER_OF'}]->(c:Community)
-            RETURN c.label AS label, c.heuristicLabel AS heuristicLabel
-            LIMIT 1
+                    const clusterQuery = `
+            MATCH (n {id: '${result.nodeId.replace(/'/g, "''")}'})-[:CodeRelation {type: 'MEMBER_OF'}]->(c:Community)
+            RETURN c.label AS label, c.heuristicLabel AS heuristicLabel
+            LIMIT 1
           `;
                     const clusters = await executeQuery(repo.id, clusterQuery);
                     if (clusters.length > 0) {
@@ -294,10 +254,10 @@ export class LocalBackend {
             // Add relationships if depth is 'full' and we have a node ID
             if (depth === 'full' && result.nodeId) {
                 try {
-                    const relQuery = `
-            MATCH (n {id: '${result.nodeId.replace(/'/g, "''")}'})-[r:CodeRelation]->(m)
-            RETURN r.type AS type, m.name AS targetName, m.filePath AS targetPath
-            LIMIT 5
+                    const relQuery = `
+            MATCH (n {id: '${result.nodeId.replace(/'/g, "''")}'})-[r:CodeRelation]->(m)
+            RETURN r.type AS type, m.name AS targetName, m.filePath AS targetPath
+            LIMIT 5
           `;
                     const rels = await executeQuery(repo.id, relQuery);
                     result.connections = rels.map((rel) => ({
@@ -324,11 +284,11 @@ export class LocalBackend {
         for (const bm25Result of bm25Results) {
             const fileName = bm25Result.filePath.split('/').pop() || bm25Result.filePath;
             try {
-                const symbolQuery = `
-          MATCH (n) 
-          WHERE n.filePath CONTAINS '${fileName.replace(/'/g, "''")}'
-          RETURN n.id AS id, n.name AS name, labels(n)[0] AS type, n.filePath AS filePath, n.startLine AS startLine, n.endLine AS endLine
-          LIMIT 3
+                const symbolQuery = `
+          MATCH (n) 
+          WHERE n.filePath CONTAINS '${fileName.replace(/'/g, "''")}'
+          RETURN n.id AS id, n.name AS name, labels(n)[0] AS type, n.filePath AS filePath, n.startLine AS startLine, n.endLine AS endLine
+          LIMIT 3
         `;
                 const symbols = await executeQuery(repo.id, symbolQuery);
                 if (symbols.length > 0) {
@@ -372,14 +332,14 @@ export class LocalBackend {
             const queryVec = await embedQuery(query);
             const dims = getEmbeddingDims();
             const queryVecStr = `[${queryVec.join(',')}]`;
-            const vectorQuery = `
-        CALL QUERY_VECTOR_INDEX('CodeEmbedding', 'code_embedding_idx', 
-          CAST(${queryVecStr} AS FLOAT[${dims}]), ${limit})
-        YIELD node AS emb, distance
-        WITH emb, distance
-        WHERE distance < 0.6
-        RETURN emb.nodeId AS nodeId, distance
-        ORDER BY distance
+            const vectorQuery = `
+        CALL QUERY_VECTOR_INDEX('CodeEmbedding', 'code_embedding_idx', 
+          CAST(${queryVecStr} AS FLOAT[${dims}]), ${limit})
+        YIELD node AS emb, distance
+        WITH emb, distance
+        WHERE distance < 0.6
+        RETURN emb.nodeId AS nodeId, distance
+        ORDER BY distance
       `;
             const embResults = await executeQuery(repo.id, vectorQuery);
             if (embResults.length === 0)
@@ -430,6 +390,42 @@ export class LocalBackend {
             return { error: err.message || 'Query failed' };
         }
     }
+    /**
+     * Aggregate same-named clusters: group by heuristicLabel, sum symbols,
+     * weighted-average cohesion, filter out tiny clusters (<5 symbols).
+     * Raw communities stay intact in KuzuDB for Cypher queries.
+     */
+    aggregateClusters(clusters) {
+        const groups = new Map();
+        for (const c of clusters) {
+            const label = c.heuristicLabel || c.label || 'Unknown';
+            const symbols = c.symbolCount || 0;
+            const cohesion = c.cohesion || 0;
+            const existing = groups.get(label);
+            if (!existing) {
+                groups.set(label, { ids: [c.id], totalSymbols: symbols, weightedCohesion: cohesion * symbols, largest: c });
+            }
+            else {
+                existing.ids.push(c.id);
+                existing.totalSymbols += symbols;
+                existing.weightedCohesion += cohesion * symbols;
+                if (symbols > (existing.largest.symbolCount || 0)) {
+                    existing.largest = c;
+                }
+            }
+        }
+        return Array.from(groups.entries())
+            .map(([label, g]) => ({
+            id: g.largest.id,
+            label,
+            heuristicLabel: label,
+            symbolCount: g.totalSymbols,
+            cohesion: g.totalSymbols > 0 ? g.weightedCohesion / g.totalSymbols : 0,
+            subCommunities: g.ids.length,
+        }))
+            .filter(c => c.symbolCount >= 5)
+            .sort((a, b) => b.symbolCount - a.symbolCount);
+    }
     async overview(repo, params) {
         await this.ensureInitialized(repo.id);
         const limit = params.limit || 20;
@@ -442,19 +438,22 @@ export class LocalBackend {
         };
         if (params.showClusters !== false) {
             try {
-                const clusters = await executeQuery(repo.id, `
-          MATCH (c:Community)
-          RETURN c.id AS id, c.label AS label, c.heuristicLabel AS heuristicLabel, c.cohesion AS cohesion, c.symbolCount AS symbolCount
-          ORDER BY c.symbolCount DESC
-          LIMIT ${limit}
+                // Fetch more raw communities than the display limit so aggregation has enough data
+                const rawLimit = Math.max(limit * 5, 200);
+                const clusters = await executeQuery(repo.id, `
+          MATCH (c:Community)
+          RETURN c.id AS id, c.label AS label, c.heuristicLabel AS heuristicLabel, c.cohesion AS cohesion, c.symbolCount AS symbolCount
+          ORDER BY c.symbolCount DESC
+          LIMIT ${rawLimit}
         `);
-                result.clusters = clusters.map((c) => ({
+                const rawClusters = clusters.map((c) => ({
                     id: c.id || c[0],
                     label: c.label || c[1],
                     heuristicLabel: c.heuristicLabel || c[2],
                     cohesion: c.cohesion || c[3],
                     symbolCount: c.symbolCount || c[4],
                 }));
+                result.clusters = this.aggregateClusters(rawClusters).slice(0, limit);
             }
             catch {
                 result.clusters = [];
@@ -462,11 +461,11 @@ export class LocalBackend {
         }
         if (params.showProcesses !== false) {
             try {
-                const processes = await executeQuery(repo.id, `
-          MATCH (p:Process)
-          RETURN p.id AS id, p.label AS label, p.heuristicLabel AS heuristicLabel, p.processType AS processType, p.stepCount AS stepCount
-          ORDER BY p.stepCount DESC
-          LIMIT ${limit}
+                const processes = await executeQuery(repo.id, `
+          MATCH (p:Process)
+          RETURN p.id AS id, p.label AS label, p.heuristicLabel AS heuristicLabel, p.processType AS processType, p.stepCount AS stepCount
+          ORDER BY p.stepCount DESC
+          LIMIT ${limit}
         `);
                 result.processes = processes.map((p) => ({
                     id: p.id || p[0],
@@ -486,33 +485,33 @@ export class LocalBackend {
         await this.ensureInitialized(repo.id);
         const { name, type } = params;
         if (type === 'symbol') {
-            const symbolQuery = `
-        MATCH (n)
-        WHERE n.name = '${name.replace(/'/g, "''")}'
-        RETURN n.id AS id, n.name AS name, labels(n)[0] AS type, n.filePath AS filePath, n.startLine AS startLine, n.endLine AS endLine
-        LIMIT 1
+            const symbolQuery = `
+        MATCH (n)
+        WHERE n.name = '${name.replace(/'/g, "''")}'
+        RETURN n.id AS id, n.name AS name, labels(n)[0] AS type, n.filePath AS filePath, n.startLine AS startLine, n.endLine AS endLine
+        LIMIT 1
       `;
             const symbols = await executeQuery(repo.id, symbolQuery);
             if (symbols.length === 0)
                 return { error: `Symbol '${name}' not found` };
             const sym = symbols[0];
             const symId = sym.id || sym[0];
-            const callersQuery = `
-        MATCH (caller)-[:CodeRelation {type: 'CALLS'}]->(n {id: '${symId}'})
-        RETURN caller.name AS name, caller.filePath AS filePath
-        LIMIT 10
+            const callersQuery = `
+        MATCH (caller)-[:CodeRelation {type: 'CALLS'}]->(n {id: '${symId}'})
+        RETURN caller.name AS name, caller.filePath AS filePath
+        LIMIT 10
       `;
             const callers = await executeQuery(repo.id, callersQuery);
-            const calleesQuery = `
-        MATCH (n {id: '${symId}'})-[:CodeRelation {type: 'CALLS'}]->(callee)
-        RETURN callee.name AS name, callee.filePath AS filePath
-        LIMIT 10
+            const calleesQuery = `
+        MATCH (n {id: '${symId}'})-[:CodeRelation {type: 'CALLS'}]->(callee)
+        RETURN callee.name AS name, callee.filePath AS filePath
+        LIMIT 10
       `;
             const callees = await executeQuery(repo.id, calleesQuery);
-            const communityQuery = `
-        MATCH (n {id: '${symId}'})-[:CodeRelation {type: 'MEMBER_OF'}]->(c:Community)
-        RETURN c.label AS label, c.heuristicLabel AS heuristicLabel
-        LIMIT 1
+            const communityQuery = `
+        MATCH (n {id: '${symId}'})-[:CodeRelation {type: 'MEMBER_OF'}]->(c:Community)
+        RETURN c.label AS label, c.heuristicLabel AS heuristicLabel
+        LIMIT 1
       `;
             const communities = await executeQuery(repo.id, communityQuery);
             return {
@@ -533,30 +532,47 @@ export class LocalBackend {
             };
         }
         if (type === 'cluster') {
-            const clusterQuery = `
-        MATCH (c:Community)
-        WHERE c.label = '${name.replace(/'/g, "''")}' OR c.heuristicLabel = '${name.replace(/'/g, "''")}'
-        RETURN c.id AS id, c.label AS label, c.heuristicLabel AS heuristicLabel, c.cohesion AS cohesion, c.symbolCount AS symbolCount
-        LIMIT 1
+            const escaped = name.replace(/'/g, "''");
+            // Find ALL communities with this label (not just one)
+            const clusterQuery = `
+        MATCH (c:Community)
+        WHERE c.label = '${escaped}' OR c.heuristicLabel = '${escaped}'
+        RETURN c.id AS id, c.label AS label, c.heuristicLabel AS heuristicLabel, c.cohesion AS cohesion, c.symbolCount AS symbolCount
       `;
             const clusters = await executeQuery(repo.id, clusterQuery);
             if (clusters.length === 0)
                 return { error: `Cluster '${name}' not found` };
-            const cluster = clusters[0];
-            const clusterId = cluster.id || cluster[0];
-            const membersQuery = `
-        MATCH (n)-[:CodeRelation {type: 'MEMBER_OF'}]->(c {id: '${clusterId}'})
-        RETURN n.name AS name, labels(n)[0] AS type, n.filePath AS filePath
-        LIMIT 20
+            const rawClusters = clusters.map((c) => ({
+                id: c.id || c[0],
+                label: c.label || c[1],
+                heuristicLabel: c.heuristicLabel || c[2],
+                cohesion: c.cohesion || c[3],
+                symbolCount: c.symbolCount || c[4],
+            }));
+            // Aggregate: sum symbols, weighted-average cohesion across sub-communities
+            let totalSymbols = 0;
+            let weightedCohesion = 0;
+            for (const c of rawClusters) {
+                const s = c.symbolCount || 0;
+                totalSymbols += s;
+                weightedCohesion += (c.cohesion || 0) * s;
+            }
+            // Fetch members from ALL matching sub-communities (DISTINCT to avoid dupes)
+            const membersQuery = `
+        MATCH (n)-[:CodeRelation {type: 'MEMBER_OF'}]->(c:Community)
+        WHERE c.label = '${escaped}' OR c.heuristicLabel = '${escaped}'
+        RETURN DISTINCT n.name AS name, labels(n)[0] AS type, n.filePath AS filePath
+        LIMIT 30
       `;
             const members = await executeQuery(repo.id, membersQuery);
             return {
                 cluster: {
-                    id: clusterId,
-                    label: cluster.label || cluster[1],
-                    heuristicLabel: cluster.heuristicLabel || cluster[2],
-                    cohesion: cluster.cohesion || cluster[3],
-                    symbolCount: cluster.symbolCount || cluster[4],
+                    id: rawClusters[0].id,
+                    label: rawClusters[0].heuristicLabel || rawClusters[0].label,
+                    heuristicLabel: rawClusters[0].heuristicLabel || rawClusters[0].label,
+                    cohesion: totalSymbols > 0 ? weightedCohesion / totalSymbols : 0,
+                    symbolCount: totalSymbols,
+                    subCommunities: rawClusters.length,
                 },
                 members: members.map((m) => ({
                     name: m.name || m[0],
@@ -566,21 +582,21 @@ export class LocalBackend {
             };
         }
         if (type === 'process') {
-            const processQuery = `
-        MATCH (p:Process)
-        WHERE p.label = '${name.replace(/'/g, "''")}' OR p.heuristicLabel = '${name.replace(/'/g, "''")}'
-        RETURN p.id AS id, p.label AS label, p.heuristicLabel AS heuristicLabel, p.processType AS processType, p.stepCount AS stepCount, p.entryPointId AS entryPointId, p.terminalId AS terminalId
-        LIMIT 1
+            const processQuery = `
+        MATCH (p:Process)
+        WHERE p.label = '${name.replace(/'/g, "''")}' OR p.heuristicLabel = '${name.replace(/'/g, "''")}'
+        RETURN p.id AS id, p.label AS label, p.heuristicLabel AS heuristicLabel, p.processType AS processType, p.stepCount AS stepCount, p.entryPointId AS entryPointId, p.terminalId AS terminalId
+        LIMIT 1
       `;
             const processes = await executeQuery(repo.id, processQuery);
             if (processes.length === 0)
                 return { error: `Process '${name}' not found` };
             const proc = processes[0];
             const procId = proc.id || proc[0];
-            const stepsQuery = `
-        MATCH (n)-[r:CodeRelation {type: 'STEP_IN_PROCESS'}]->(p {id: '${procId}'})
-        RETURN n.name AS name, labels(n)[0] AS type, n.filePath AS filePath, r.step AS step
-        ORDER BY r.step
+            const stepsQuery = `
+        MATCH (n)-[r:CodeRelation {type: 'STEP_IN_PROCESS'}]->(p {id: '${procId}'})
+        RETURN n.name AS name, labels(n)[0] AS type, n.filePath AS filePath, r.step AS step
+        ORDER BY r.step
       `;
             const steps = await executeQuery(repo.id, stepsQuery);
             return {
@@ -612,11 +628,11 @@ export class LocalBackend {
         const minConfidence = params.minConfidence ?? 0;
         const relTypeFilter = relationTypes.map(t => `'${t}'`).join(', ');
         const confidenceFilter = minConfidence > 0 ? ` AND r.confidence >= ${minConfidence}` : '';
-        const targetQuery = `
-      MATCH (n)
-      WHERE n.name = '${target.replace(/'/g, "''")}'
-      RETURN n.id AS id, n.name AS name, labels(n)[0] AS type, n.filePath AS filePath
-      LIMIT 1
+        const targetQuery = `
+      MATCH (n)
+      WHERE n.name = '${target.replace(/'/g, "''")}'
+      RETURN n.id AS id, n.name AS name, labels(n)[0] AS type, n.filePath AS filePath
+      LIMIT 1
     `;
         const targets = await executeQuery(repo.id, targetQuery);
         if (targets.length === 0)
@@ -772,14 +788,9 @@ export class LocalBackend {
                 stats: {
                     fileCount: newMeta.stats.files || 0,
                     functionCount: newMeta.stats.nodes || 0,
-                    classCount: 0,
-                    interfaceCount: 0,
-                    methodCount: 0,
                     communityCount: newMeta.stats.communities || 0,
                     processCount: newMeta.stats.processes || 0,
                 },
-                hotspots: [],
-                folderTree: '',
             });
             console.error('GitNexus: Indexing complete!');
             return {

package/dist/mcp/resources.d.ts

@@ -2,8 +2,7 @@
  * MCP Resources (Multi-Repo)
  *
  * Provides structured on-demand data to AI agents.
- * Supports repo-scoped URIs (gitnexus://repo/{name}/context)
- * and backwards-compatible global URIs (gitnexus://context).
+ * All resources use repo-scoped URIs: gitnexus://repo/{name}/context
  */
 import type { LocalBackend } from './local/local-backend.js';
 export interface ResourceDefinition {