@ruvector/ruvllm 0.2.3 → 2.4.0

package/scripts/hybrid-model-compare.js

@@ -0,0 +1,373 @@
+#!/usr/bin/env node
+/**
+ * Hybrid Model Comparison
+ *
+ * Combines embedding similarity with keyword boosting.
+ * This addresses the "reviewer overfit" problem by:
+ * 1. Computing embedding similarity
+ * 2. Boosting agents that have keyword matches in the task
+ * 3. Using weighted combination for final score
+ */
+
+const { execSync } = require('child_process');
+const { existsSync } = require('fs');
+const { join } = require('path');
+const { homedir } = require('os');
+
+const MODELS_DIR = join(homedir(), '.ruvllm', 'models');
+const RUVLTRA_MODEL = join(MODELS_DIR, 'ruvltra-claude-code-0.5b-q4_k_m.gguf');
+const QWEN_MODEL = join(MODELS_DIR, 'qwen2.5-0.5b-instruct-q4_k_m.gguf');
+
+// V1 descriptions for embedding
+const DESCRIPTIONS_V1 = {
+  coder: 'implement create write build add code function class component feature',
+  researcher: 'research find investigate analyze explore search discover examine',
+  reviewer: 'review check evaluate assess inspect examine code quality',
+  tester: 'test unit integration e2e coverage mock assertion spec',
+  architect: 'design architecture schema system structure plan database',
+  'security-architect': 'security vulnerability xss injection audit cve authentication',
+  debugger: 'debug fix bug error issue broken crash exception trace',
+  documenter: 'document readme jsdoc comment explain describe documentation',
+  refactorer: 'refactor extract rename consolidate clean restructure simplify',
+  optimizer: 'optimize performance slow fast cache speed memory latency',
+  devops: 'deploy ci cd kubernetes docker pipeline container infrastructure',
+  'api-docs': 'openapi swagger api documentation graphql schema endpoint',
+  planner: 'plan estimate prioritize sprint roadmap schedule milestone',
+};
+
+// UNIQUE trigger keywords - words that strongly indicate a specific agent
+// Priority-ordered: first match wins for disambiguation
+// NOTE: "investigate" takes priority over "slow" for researcher vs optimizer
+const TRIGGER_KEYWORDS = {
+  // Higher priority agents (check these first)
+  researcher: ['research', 'investigate', 'explore', 'discover', 'best practices', 'patterns', 'analyze', 'look into', 'find out'],
+  coder: ['implement', 'build', 'create', 'component', 'function', 'typescript', 'react', 'feature', 'write code'],
+  tester: ['test', 'tests', 'testing', 'unit test', 'integration test', 'e2e', 'coverage', 'spec'],
+  reviewer: ['review', 'pull request', 'pr', 'code quality', 'code review', 'check code'],
+  debugger: ['debug', 'fix', 'bug', 'error', 'exception', 'crash', 'trace', 'null pointer', 'memory leak'],
+  'security-architect': ['security', 'vulnerability', 'xss', 'injection', 'csrf', 'cve', 'audit', 'exploit'],
+  refactorer: ['refactor', 'async/await', 'modernize', 'restructure', 'extract', 'legacy'],
+  // Optimizer: removed "slow" (too generic), added query-specific terms
+  optimizer: ['optimize', 'performance', 'cache', 'caching', 'speed up', 'latency', 'faster', 'queries', 'reduce time'],
+  architect: ['design', 'architecture', 'schema', 'structure', 'diagram', 'system design', 'plan architecture'],
+  documenter: ['jsdoc', 'comment', 'comments', 'readme', 'documentation', 'document', 'explain'],
+  devops: ['deploy', 'ci/cd', 'kubernetes', 'docker', 'pipeline', 'infrastructure', 'container'],
+  'api-docs': ['openapi', 'swagger', 'api doc', 'rest api', 'graphql', 'endpoint'],
+  planner: ['sprint', 'plan', 'roadmap', 'milestone', 'estimate', 'schedule', 'prioritize'],
+};
+
+// Priority order for disambiguation (when multiple agents match)
+const AGENT_PRIORITY = [
+  'researcher',     // "investigate" wins over "slow"
+  'debugger',       // "fix" wins over generic terms
+  'tester',         // "test" is specific
+  'security-architect',
+  'coder',
+  'reviewer',
+  'refactorer',
+  'optimizer',
+  'architect',
+  'documenter',
+  'devops',
+  'api-docs',
+  'planner',
+];
+
+const ROUTING_TESTS = [
+  { task: 'Implement a binary search function in TypeScript', expected: 'coder' },
+  { task: 'Write unit tests for the authentication module', expected: 'tester' },
+  { task: 'Review the pull request for security vulnerabilities', expected: 'reviewer' },
+  { task: 'Research best practices for React state management', expected: 'researcher' },
+  { task: 'Design the database schema for user profiles', expected: 'architect' },
+  { task: 'Fix the null pointer exception in the login handler', expected: 'debugger' },
+  { task: 'Audit the API endpoints for XSS vulnerabilities', expected: 'security-architect' },
+  { task: 'Write JSDoc comments for the utility functions', expected: 'documenter' },
+  { task: 'Refactor the payment module to use async/await', expected: 'refactorer' },
+  { task: 'Optimize the database queries for the dashboard', expected: 'optimizer' },
+  { task: 'Set up the CI/CD pipeline for the microservices', expected: 'devops' },
+  { task: 'Generate OpenAPI documentation for the REST API', expected: 'api-docs' },
+  { task: 'Create a sprint plan for the next two weeks', expected: 'planner' },
+  { task: 'Build a React component for user registration', expected: 'coder' },
+  { task: 'Debug memory leak in the WebSocket handler', expected: 'debugger' },
+  { task: 'Investigate slow API response times', expected: 'researcher' },
+  { task: 'Check code for potential race conditions', expected: 'reviewer' },
+  { task: 'Add integration tests for the payment gateway', expected: 'tester' },
+  { task: 'Plan the architecture for real-time notifications', expected: 'architect' },
+  { task: 'Cache the frequently accessed user data', expected: 'optimizer' },
+];
+
+function getEmbedding(modelPath, text) {
+  try {
+    const sanitized = text.replace(/"/g, '\\"').replace(/\n/g, ' ');
+    const result = execSync(
+      `llama-embedding -m "${modelPath}" -p "${sanitized}" --embd-output-format json 2>/dev/null`,
+      { encoding: 'utf-8', maxBuffer: 10 * 1024 * 1024 }
+    );
+    const json = JSON.parse(result);
+    return json.data[json.data.length - 1].embedding;
+  } catch {
+    return null;
+  }
+}
+
+function cosineSimilarity(a, b) {
+  if (!a || !b || a.length !== b.length) return 0;
+  let dot = 0, normA = 0, normB = 0;
+  for (let i = 0; i < a.length; i++) {
+    dot += a[i] * b[i];
+    normA += a[i] * a[i];
+    normB += b[i] * b[i];
+  }
+  return dot / (Math.sqrt(normA) * Math.sqrt(normB) || 1);
+}
+
+/**
+ * Count keyword matches for each agent
+ */
+function getKeywordScores(task) {
+  const taskLower = task.toLowerCase();
+  const scores = {};
+
+  for (const [agent, keywords] of Object.entries(TRIGGER_KEYWORDS)) {
+    let matches = 0;
+    for (const kw of keywords) {
+      if (taskLower.includes(kw.toLowerCase())) {
+        matches++;
+      }
+    }
+    scores[agent] = matches;
+  }
+
+  return scores;
+}
+
+/**
+ * Pure embedding routing (baseline)
+ */
+function routeEmbeddingOnly(taskEmbedding, agentEmbeddings) {
+  let bestAgent = 'coder';
+  let bestSim = -1;
+
+  for (const [agent, emb] of Object.entries(agentEmbeddings)) {
+    const sim = cosineSimilarity(taskEmbedding, emb);
+    if (sim > bestSim) {
+      bestSim = sim;
+      bestAgent = agent;
+    }
+  }
+
+  return { agent: bestAgent, confidence: bestSim };
+}
+
+/**
+ * Pure keyword routing
+ */
+function routeKeywordOnly(task) {
+  const scores = getKeywordScores(task);
+  let bestAgent = 'coder';
+  let bestScore = 0;
+
+  for (const [agent, score] of Object.entries(scores)) {
+    if (score > bestScore) {
+      bestScore = score;
+      bestAgent = agent;
+    }
+  }
+
+  return { agent: bestAgent, confidence: bestScore };
+}
+
+/**
+ * Hybrid routing - combine embedding similarity with keyword boost
+ */
+function routeHybrid(task, taskEmbedding, agentEmbeddings, embeddingWeight = 0.6, keywordWeight = 0.4) {
+  const keywordScores = getKeywordScores(task);
+
+  // Normalize keyword scores to 0-1 range
+  const maxKeyword = Math.max(...Object.values(keywordScores), 1);
+  const normalizedKeywords = {};
+  for (const agent of Object.keys(keywordScores)) {
+    normalizedKeywords[agent] = keywordScores[agent] / maxKeyword;
+  }
+
+  let bestAgent = 'coder';
+  let bestScore = -1;
+  const allScores = {};
+
+  for (const [agent, emb] of Object.entries(agentEmbeddings)) {
+    const embSim = cosineSimilarity(taskEmbedding, emb);
+    const kwScore = normalizedKeywords[agent] || 0;
+    const combined = embeddingWeight * embSim + keywordWeight * kwScore;
+    allScores[agent] = { embedding: embSim, keyword: kwScore, combined };
+
+    if (combined > bestScore) {
+      bestScore = combined;
+      bestAgent = agent;
+    }
+  }
+
+  return { agent: bestAgent, confidence: bestScore, scores: allScores };
+}
+
+/**
+ * Keyword-first routing - use keywords as primary, embedding as tiebreaker
+ */
+function routeKeywordFirst(task, taskEmbedding, agentEmbeddings) {
+  const keywordScores = getKeywordScores(task);
+
+  // Find agents with max keyword matches
+  const maxKw = Math.max(...Object.values(keywordScores));
+
+  if (maxKw > 0) {
+    // At least one keyword match - use keywords, embedding as tiebreaker
+    const candidates = Object.entries(keywordScores)
+      .filter(([_, score]) => score === maxKw)
+      .map(([agent, _]) => agent);
+
+    if (candidates.length === 1) {
+      return { agent: candidates[0], confidence: maxKw };
+    }
+
+    // Multiple candidates with same keyword count - use embedding
+    let bestAgent = candidates[0];
+    let bestSim = -1;
+    for (const agent of candidates) {
+      const sim = cosineSimilarity(taskEmbedding, agentEmbeddings[agent]);
+      if (sim > bestSim) {
+        bestSim = sim;
+        bestAgent = agent;
+      }
+    }
+    return { agent: bestAgent, confidence: maxKw + bestSim / 10 };
+  }
+
+  // No keyword matches - fall back to pure embedding
+  return routeEmbeddingOnly(taskEmbedding, agentEmbeddings);
+}
+
+function runBenchmark(modelPath, routerFn, name) {
+  const agentEmbeddings = {};
+  for (const [agent, desc] of Object.entries(DESCRIPTIONS_V1)) {
+    agentEmbeddings[agent] = getEmbedding(modelPath, desc);
+  }
+
+  let correct = 0;
+  const results = [];
+
+  for (const test of ROUTING_TESTS) {
+    const taskEmb = getEmbedding(modelPath, test.task);
+    const { agent } = routerFn(test.task, taskEmb, agentEmbeddings);
+    const isCorrect = agent === test.expected;
+    if (isCorrect) correct++;
+    results.push({ task: test.task, expected: test.expected, got: agent, correct: isCorrect });
+  }
+
+  return { accuracy: correct / ROUTING_TESTS.length, correct, total: ROUTING_TESTS.length, results, name };
+}
+
+async function main() {
+  console.log('\n╔═══════════════════════════════════════════════════════════════════════════════════╗');
+  console.log('║                 HYBRID ROUTING: Embeddings + Keywords                             ║');
+  console.log('╚═══════════════════════════════════════════════════════════════════════════════════╝\n');
+
+  if (!existsSync(RUVLTRA_MODEL)) {
+    console.error('RuvLTRA model not found.');
+    process.exit(1);
+  }
+
+  console.log('Strategies:');
+  console.log('  1. Embedding Only (baseline)');
+  console.log('  2. Keyword Only (no model)');
+  console.log('  3. Hybrid 60/40 (60% embedding, 40% keyword)');
+  console.log('  4. Hybrid 40/60 (40% embedding, 60% keyword)');
+  console.log('  5. Keyword-First (keywords primary, embedding tiebreaker)\n');
+
+  // RuvLTRA tests
+  console.log('─────────────────────────────────────────────────────────────────');
+  console.log('                       RUVLTRA RESULTS');
+  console.log('─────────────────────────────────────────────────────────────────\n');
+
+  const ruvEmbedding = runBenchmark(RUVLTRA_MODEL,
+    (task, taskEmb, agentEmbs) => routeEmbeddingOnly(taskEmb, agentEmbs),
+    'Embedding Only');
+  console.log(`  Embedding Only:   ${(ruvEmbedding.accuracy * 100).toFixed(1)}%`);
+
+  const ruvKeyword = runBenchmark(RUVLTRA_MODEL,
+    (task, taskEmb, agentEmbs) => routeKeywordOnly(task),
+    'Keyword Only');
+  console.log(`  Keyword Only:     ${(ruvKeyword.accuracy * 100).toFixed(1)}%`);
+
+  const ruvHybrid60 = runBenchmark(RUVLTRA_MODEL,
+    (task, taskEmb, agentEmbs) => routeHybrid(task, taskEmb, agentEmbs, 0.6, 0.4),
+    'Hybrid 60/40');
+  console.log(`  Hybrid 60/40:     ${(ruvHybrid60.accuracy * 100).toFixed(1)}%`);
+
+  const ruvHybrid40 = runBenchmark(RUVLTRA_MODEL,
+    (task, taskEmb, agentEmbs) => routeHybrid(task, taskEmb, agentEmbs, 0.4, 0.6),
+    'Hybrid 40/60');
+  console.log(`  Hybrid 40/60:     ${(ruvHybrid40.accuracy * 100).toFixed(1)}%`);
+
+  const ruvKwFirst = runBenchmark(RUVLTRA_MODEL,
+    (task, taskEmb, agentEmbs) => routeKeywordFirst(task, taskEmb, agentEmbs),
+    'Keyword-First');
+  console.log(`  Keyword-First:    ${(ruvKwFirst.accuracy * 100).toFixed(1)}%`);
+
+  // Qwen tests
+  console.log('\n─────────────────────────────────────────────────────────────────');
+  console.log('                       QWEN RESULTS');
+  console.log('─────────────────────────────────────────────────────────────────\n');
+
+  const qwenEmbedding = runBenchmark(QWEN_MODEL,
+    (task, taskEmb, agentEmbs) => routeEmbeddingOnly(taskEmb, agentEmbs),
+    'Embedding Only');
+  console.log(`  Embedding Only:   ${(qwenEmbedding.accuracy * 100).toFixed(1)}%`);
+
+  const qwenHybrid60 = runBenchmark(QWEN_MODEL,
+    (task, taskEmb, agentEmbs) => routeHybrid(task, taskEmb, agentEmbs, 0.6, 0.4),
+    'Hybrid 60/40');
+  console.log(`  Hybrid 60/40:     ${(qwenHybrid60.accuracy * 100).toFixed(1)}%`);
+
+  const qwenKwFirst = runBenchmark(QWEN_MODEL,
+    (task, taskEmb, agentEmbs) => routeKeywordFirst(task, taskEmb, agentEmbs),
+    'Keyword-First');
+  console.log(`  Keyword-First:    ${(qwenKwFirst.accuracy * 100).toFixed(1)}%`);
+
+  // Summary table
+  console.log('\n═══════════════════════════════════════════════════════════════════════════════════');
+  console.log('                              SUMMARY');
+  console.log('═══════════════════════════════════════════════════════════════════════════════════\n');
+
+  const fmt = (v) => `${(v * 100).toFixed(1)}%`.padStart(8);
+
+  console.log('┌───────────────────────┬──────────┬──────────┬──────────────────┐');
+  console.log('│ Strategy              │  RuvLTRA │   Qwen   │ RuvLTRA vs Qwen  │');
+  console.log('├───────────────────────┼──────────┼──────────┼──────────────────┤');
+  console.log(`│ Embedding Only        │${fmt(ruvEmbedding.accuracy)}  │${fmt(qwenEmbedding.accuracy)}  │     +${((ruvEmbedding.accuracy - qwenEmbedding.accuracy) * 100).toFixed(1)} pts     │`);
+  console.log(`│ Keyword Only          │${fmt(ruvKeyword.accuracy)}  │${fmt(ruvKeyword.accuracy)}  │       same       │`);
+  console.log(`│ Hybrid 60/40          │${fmt(ruvHybrid60.accuracy)}  │${fmt(qwenHybrid60.accuracy)}  │     +${((ruvHybrid60.accuracy - qwenHybrid60.accuracy) * 100).toFixed(1)} pts     │`);
+  console.log(`│ Keyword-First         │${fmt(ruvKwFirst.accuracy)}  │${fmt(qwenKwFirst.accuracy)}  │     +${((ruvKwFirst.accuracy - qwenKwFirst.accuracy) * 100).toFixed(1)} pts     │`);
+  console.log('└───────────────────────┴──────────┴──────────┴──────────────────┘');
+
+  // Best results
+  const ruvBest = [ruvEmbedding, ruvKeyword, ruvHybrid60, ruvHybrid40, ruvKwFirst]
+    .reduce((a, b) => a.accuracy > b.accuracy ? a : b);
+
+  console.log(`\n  BEST RuvLTRA: ${ruvBest.name} = ${(ruvBest.accuracy * 100).toFixed(1)}%`);
+  console.log(`  Improvement over embedding-only: +${((ruvBest.accuracy - ruvEmbedding.accuracy) * 100).toFixed(1)} points`);
+
+  // Show best results details
+  console.log('\n─────────────────────────────────────────────────────────────────');
+  console.log(`               BEST STRATEGY DETAILS: ${ruvBest.name}`);
+  console.log('─────────────────────────────────────────────────────────────────\n');
+
+  for (const r of ruvBest.results) {
+    const mark = r.correct ? '✓' : '✗';
+    const task = r.task.slice(0, 45).padEnd(45);
+    const exp = r.expected.padEnd(18);
+    console.log(`${mark} ${task} ${exp}${r.correct ? '' : '→ ' + r.got}`);
+  }
+
+  console.log('\n');
+}
+
+main().catch(console.error);

package/scripts/improved-model-compare.js

@@ -0,0 +1,288 @@
+#!/usr/bin/env node
+/**
+ * Improved Model Comparison - Enhanced Agent Descriptions
+ *
+ * Key improvements:
+ * 1. Semantic sentence descriptions instead of keyword lists
+ * 2. Example tasks embedded in descriptions
+ * 3. Unique discriminating phrases for each agent
+ * 4. Adjusted similarity scoring with top-k voting
+ */
+
+const { execSync } = require('child_process');
+const { existsSync } = require('fs');
+const { join } = require('path');
+const { homedir } = require('os');
+
+// Model paths
+const MODELS_DIR = join(homedir(), '.ruvllm', 'models');
+const QWEN_MODEL = join(MODELS_DIR, 'qwen2.5-0.5b-instruct-q4_k_m.gguf');
+const RUVLTRA_MODEL = join(MODELS_DIR, 'ruvltra-claude-code-0.5b-q4_k_m.gguf');
+
+// IMPROVED: Semantic sentence descriptions with examples
+const AGENT_DESCRIPTIONS_V1 = {
+  coder: 'implement create write build add code function class component feature',
+  researcher: 'research find investigate analyze explore search discover examine',
+  reviewer: 'review check evaluate assess inspect examine code quality',
+  tester: 'test unit integration e2e coverage mock assertion spec',
+  architect: 'design architecture schema system structure plan database',
+  'security-architect': 'security vulnerability xss injection audit cve authentication',
+  debugger: 'debug fix bug error issue broken crash exception trace',
+  documenter: 'document readme jsdoc comment explain describe documentation',
+  refactorer: 'refactor extract rename consolidate clean restructure simplify',
+  optimizer: 'optimize performance slow fast cache speed memory latency',
+  devops: 'deploy ci cd kubernetes docker pipeline container infrastructure',
+  'api-docs': 'openapi swagger api documentation graphql schema endpoint',
+  planner: 'plan estimate prioritize sprint roadmap schedule milestone',
+};
+
+// V2: Semantic sentences with task context
+const AGENT_DESCRIPTIONS_V2 = {
+  coder: 'I write new code and implement features. Create functions, build components, implement algorithms like binary search, build React components, write TypeScript code.',
+  researcher: 'I research and investigate topics. Find best practices, explore solutions, investigate performance issues, analyze patterns, discover new approaches.',
+  reviewer: 'I review existing code for quality. Check pull requests, evaluate code style, assess readability, inspect for bugs, examine code patterns.',
+  tester: 'I write tests for code. Create unit tests, add integration tests, write e2e tests, mock dependencies, check test coverage, write test specs.',
+  architect: 'I design system architecture. Plan database schemas, design API structures, create system diagrams, plan microservices, design data models.',
+  'security-architect': 'I audit security vulnerabilities. Check for XSS, SQL injection, CSRF, audit authentication, review security policies, scan for CVEs.',
+  debugger: 'I fix bugs and debug errors. Trace exceptions, fix crashes, resolve null pointer errors, debug memory leaks, fix runtime issues.',
+  documenter: 'I write documentation and comments. Add JSDoc comments, write README files, explain code functionality, describe APIs, create guides.',
+  refactorer: 'I refactor and restructure code. Modernize to async/await, extract functions, rename variables, consolidate duplicate code, simplify logic.',
+  optimizer: 'I optimize performance and speed. Cache data, improve query performance, reduce latency, optimize memory usage, speed up slow operations.',
+  devops: 'I handle deployment and infrastructure. Set up CI/CD pipelines, configure Kubernetes, manage Docker containers, deploy to cloud.',
+  'api-docs': 'I create API documentation specs. Generate OpenAPI specs, write Swagger docs, document REST endpoints, create GraphQL schemas.',
+  planner: 'I create project plans and estimates. Sprint planning, roadmap creation, milestone tracking, task prioritization, schedule estimation.',
+};
+
+// V3: Even more specific with negative space
+const AGENT_DESCRIPTIONS_V3 = {
+  coder: 'Software developer who implements new features and writes production code. Tasks: implement binary search, build React components, create TypeScript functions, add new functionality to applications.',
+  researcher: 'Technical researcher who investigates and analyzes. Tasks: research best practices, explore state management options, investigate slow response times, analyze codebase patterns.',
+  reviewer: 'Code reviewer who evaluates existing code quality. Tasks: review pull requests, check for race conditions, assess code style, evaluate implementation approaches.',
+  tester: 'QA engineer who writes automated tests. Tasks: write unit tests, add integration tests, create e2e test suites, test payment gateways, verify authentication modules.',
+  architect: 'System architect who designs software structure. Tasks: design database schemas, plan real-time notification systems, architect microservices, model data relationships.',
+  'security-architect': 'Security specialist who audits vulnerabilities. Tasks: audit API endpoints for XSS, check SQL injection risks, review authentication security, scan for CSRF vulnerabilities.',
+  debugger: 'Bug hunter who fixes errors and traces issues. Tasks: fix null pointer exceptions, debug memory leaks, trace WebSocket errors, resolve crash bugs.',
+  documenter: 'Technical writer who creates documentation. Tasks: write JSDoc comments, create README files, document utility functions, explain complex code.',
+  refactorer: 'Code modernizer who restructures without changing behavior. Tasks: refactor to async/await, extract reusable functions, modernize legacy patterns, simplify complex logic.',
+  optimizer: 'Performance engineer who speeds up slow code. Tasks: cache frequently accessed data, optimize database queries, reduce API latency, improve memory efficiency.',
+  devops: 'DevOps engineer who manages deployment infrastructure. Tasks: set up CI/CD pipelines, configure Kubernetes clusters, manage Docker deployments, automate releases.',
+  'api-docs': 'API documentation specialist. Tasks: generate OpenAPI documentation, create Swagger specs, document REST API endpoints, write API reference guides.',
+  planner: 'Project planner who organizes work. Tasks: create sprint plans, estimate timelines, prioritize backlog, schedule milestones, plan roadmaps.',
+};
+
+// Test cases for routing
+const ROUTING_TESTS = [
+  { task: 'Implement a binary search function in TypeScript', expected: 'coder' },
+  { task: 'Write unit tests for the authentication module', expected: 'tester' },
+  { task: 'Review the pull request for security vulnerabilities', expected: 'reviewer' },
+  { task: 'Research best practices for React state management', expected: 'researcher' },
+  { task: 'Design the database schema for user profiles', expected: 'architect' },
+  { task: 'Fix the null pointer exception in the login handler', expected: 'debugger' },
+  { task: 'Audit the API endpoints for XSS vulnerabilities', expected: 'security-architect' },
+  { task: 'Write JSDoc comments for the utility functions', expected: 'documenter' },
+  { task: 'Refactor the payment module to use async/await', expected: 'refactorer' },
+  { task: 'Optimize the database queries for the dashboard', expected: 'optimizer' },
+  { task: 'Set up the CI/CD pipeline for the microservices', expected: 'devops' },
+  { task: 'Generate OpenAPI documentation for the REST API', expected: 'api-docs' },
+  { task: 'Create a sprint plan for the next two weeks', expected: 'planner' },
+  { task: 'Build a React component for user registration', expected: 'coder' },
+  { task: 'Debug memory leak in the WebSocket handler', expected: 'debugger' },
+  { task: 'Investigate slow API response times', expected: 'researcher' },
+  { task: 'Check code for potential race conditions', expected: 'reviewer' },
+  { task: 'Add integration tests for the payment gateway', expected: 'tester' },
+  { task: 'Plan the architecture for real-time notifications', expected: 'architect' },
+  { task: 'Cache the frequently accessed user data', expected: 'optimizer' },
+];
+
+// Similarity test pairs
+const SIMILARITY_TESTS = [
+  { text1: 'implement user authentication', text2: 'create login functionality', expected: 'high' },
+  { text1: 'write unit tests', text2: 'fix database bug', expected: 'low' },
+  { text1: 'optimize query performance', text2: 'improve database speed', expected: 'high' },
+  { text1: 'design system architecture', text2: 'plan software structure', expected: 'high' },
+  { text1: 'deploy to kubernetes', text2: 'analyze user behavior', expected: 'low' },
+  { text1: 'refactor legacy code', text2: 'restructure old module', expected: 'high' },
+  { text1: 'debug memory leak', text2: 'fix memory consumption issue', expected: 'high' },
+  { text1: 'document api endpoints', text2: 'write openapi spec', expected: 'high' },
+];
+
+/**
+ * Get embedding from model
+ */
+function getEmbedding(modelPath, text) {
+  try {
+    const sanitized = text.replace(/"/g, '\\"').replace(/\n/g, ' ');
+    const result = execSync(
+      `llama-embedding -m "${modelPath}" -p "${sanitized}" --embd-output-format json 2>/dev/null`,
+      { encoding: 'utf-8', maxBuffer: 10 * 1024 * 1024 }
+    );
+    const json = JSON.parse(result);
+    return json.data[json.data.length - 1].embedding;
+  } catch (err) {
+    console.error(`Error: ${err.message}`);
+    return null;
+  }
+}
+
+/**
+ * Cosine similarity
+ */
+function cosineSimilarity(a, b) {
+  if (!a || !b || a.length !== b.length) return 0;
+  let dot = 0, normA = 0, normB = 0;
+  for (let i = 0; i < a.length; i++) {
+    dot += a[i] * b[i];
+    normA += a[i] * a[i];
+    normB += b[i] * b[i];
+  }
+  return dot / (Math.sqrt(normA) * Math.sqrt(normB) || 1);
+}
+
+/**
+ * Route task with top-k analysis
+ */
+function routeTask(taskEmbedding, agentEmbeddings, topK = 3) {
+  const scores = [];
+  for (const [agent, embedding] of Object.entries(agentEmbeddings)) {
+    const sim = cosineSimilarity(taskEmbedding, embedding);
+    scores.push({ agent, similarity: sim });
+  }
+  scores.sort((a, b) => b.similarity - a.similarity);
+
+  return {
+    agent: scores[0].agent,
+    confidence: scores[0].similarity,
+    topK: scores.slice(0, topK),
+    margin: scores[0].similarity - scores[1].similarity,
+  };
+}
+
+/**
+ * Run benchmark for a specific description version
+ */
+function runBenchmark(modelPath, modelName, descriptions, version) {
+  console.log(`\n  [${version}] Computing agent embeddings...`);
+
+  const agentEmbeddings = {};
+  for (const [agent, description] of Object.entries(descriptions)) {
+    process.stdout.write(`    ${agent}... `);
+    agentEmbeddings[agent] = getEmbedding(modelPath, description);
+    console.log('done');
+  }
+
+  console.log(`  [${version}] Running routing tests...`);
+  let correct = 0;
+  const failures = [];
+
+  for (const test of ROUTING_TESTS) {
+    const taskEmbedding = getEmbedding(modelPath, test.task);
+    const { agent, confidence, topK, margin } = routeTask(taskEmbedding, agentEmbeddings);
+    const isCorrect = agent === test.expected;
+    if (isCorrect) {
+      correct++;
+    } else {
+      failures.push({
+        task: test.task,
+        expected: test.expected,
+        got: agent,
+        topK,
+        margin,
+      });
+    }
+  }
+
+  const accuracy = correct / ROUTING_TESTS.length;
+  return { accuracy, correct, total: ROUTING_TESTS.length, failures, version };
+}
+
+/**
+ * Main comparison
+ */
+async function main() {
+  console.log('\n╔═══════════════════════════════════════════════════════════════════════════════════╗');
+  console.log('║           IMPROVED MODEL COMPARISON: Testing Description Strategies               ║');
+  console.log('║                     Semantic Descriptions vs Keyword Lists                        ║');
+  console.log('╚═══════════════════════════════════════════════════════════════════════════════════╝\n');
+
+  if (!existsSync(QWEN_MODEL) || !existsSync(RUVLTRA_MODEL)) {
+    console.error('Models not found. Run the original comparison first.');
+    process.exit(1);
+  }
+
+  console.log('Testing 3 description strategies:');
+  console.log('  V1: Keyword lists (baseline)');
+  console.log('  V2: Semantic sentences with examples');
+  console.log('  V3: Task-specific descriptions with context\n');
+
+  // Test all three versions with RuvLTRA
+  console.log('─────────────────────────────────────────────────────────────────');
+  console.log('                   RUVLTRA CLAUDE CODE MODEL');
+  console.log('─────────────────────────────────────────────────────────────────');
+
+  const v1Results = runBenchmark(RUVLTRA_MODEL, 'RuvLTRA', AGENT_DESCRIPTIONS_V1, 'V1-Keywords');
+  const v2Results = runBenchmark(RUVLTRA_MODEL, 'RuvLTRA', AGENT_DESCRIPTIONS_V2, 'V2-Semantic');
+  const v3Results = runBenchmark(RUVLTRA_MODEL, 'RuvLTRA', AGENT_DESCRIPTIONS_V3, 'V3-TaskSpecific');
+
+  // Also test Qwen with best strategy
+  console.log('\n─────────────────────────────────────────────────────────────────');
+  console.log('                      QWEN 0.5B BASE MODEL');
+  console.log('─────────────────────────────────────────────────────────────────');
+
+  const qwenV1 = runBenchmark(QWEN_MODEL, 'Qwen', AGENT_DESCRIPTIONS_V1, 'V1-Keywords');
+  const qwenV3 = runBenchmark(QWEN_MODEL, 'Qwen', AGENT_DESCRIPTIONS_V3, 'V3-TaskSpecific');
+
+  // Results summary
+  console.log('\n═══════════════════════════════════════════════════════════════════════════════════');
+  console.log('                              RESULTS COMPARISON');
+  console.log('═══════════════════════════════════════════════════════════════════════════════════\n');
+
+  console.log('┌─────────────────────────┬───────────────┬───────────────┬───────────────┐');
+  console.log('│ Strategy                │ RuvLTRA       │ Qwen Base     │ Improvement   │');
+  console.log('├─────────────────────────┼───────────────┼───────────────┼───────────────┤');
+
+  const formatPct = (v) => `${(v * 100).toFixed(1)}%`.padStart(12);
+
+  console.log(`│ V1: Keywords            │${formatPct(v1Results.accuracy)}  │${formatPct(qwenV1.accuracy)}  │    baseline   │`);
+  console.log(`│ V2: Semantic            │${formatPct(v2Results.accuracy)}  │      -        │${formatPct(v2Results.accuracy - v1Results.accuracy)}  │`);
+  console.log(`│ V3: Task-Specific       │${formatPct(v3Results.accuracy)}  │${formatPct(qwenV3.accuracy)}  │${formatPct(v3Results.accuracy - v1Results.accuracy)}  │`);
+
+  console.log('└─────────────────────────┴───────────────┴───────────────┴───────────────┘');
+
+  // Find best strategy
+  const best = [v1Results, v2Results, v3Results].reduce((a, b) => a.accuracy > b.accuracy ? a : b);
+
+  console.log(`\n  BEST STRATEGY: ${best.version} with ${(best.accuracy * 100).toFixed(1)}% accuracy`);
+  console.log(`  Improvement over V1: +${((best.accuracy - v1Results.accuracy) * 100).toFixed(1)} percentage points`);
+
+  // Show remaining failures for best strategy
+  if (best.failures.length > 0) {
+    console.log(`\n  Remaining failures (${best.failures.length}):`);
+    for (const f of best.failures.slice(0, 5)) {
+      console.log(`    "${f.task.slice(0, 45)}..."`);
+      console.log(`      Expected: ${f.expected}, Got: ${f.got}`);
+      console.log(`      Top-3: ${f.topK.map(t => `${t.agent}(${(t.similarity * 100).toFixed(0)}%)`).join(', ')}`);
+    }
+  }
+
+  // RuvLTRA vs Qwen with best strategy
+  console.log('\n═══════════════════════════════════════════════════════════════════════════════════');
+  console.log('              FINAL COMPARISON (V3 Task-Specific Descriptions)');
+  console.log('═══════════════════════════════════════════════════════════════════════════════════\n');
+
+  console.log('┌─────────────────────────────┬───────────────┬───────────────┐');
+  console.log('│ Metric                      │ Qwen Base     │ RuvLTRA       │');
+  console.log('├─────────────────────────────┼───────────────┼───────────────┤');
+
+  const qwenWins = qwenV3.accuracy > v3Results.accuracy;
+  const ruvWins = v3Results.accuracy > qwenV3.accuracy;
+  console.log(`│ V3 Routing Accuracy         │${qwenWins ? '✓' : ' '}${formatPct(qwenV3.accuracy)}  │${ruvWins ? '✓' : ' '}${formatPct(v3Results.accuracy)}  │`);
+  console.log('└─────────────────────────────┴───────────────┴───────────────┘');
+
+  const winner = ruvWins ? 'RuvLTRA' : qwenWins ? 'Qwen' : 'Tie';
+  const margin = Math.abs(v3Results.accuracy - qwenV3.accuracy) * 100;
+
+  console.log(`\n  WINNER: ${winner} (${margin.toFixed(1)} point margin)`);
+  console.log('\n');
+}
+
+main().catch(console.error);