M.I.M.I.R - Multi-agent Intelligent Memory & Insight Repository

import { describe, it, expect } from 'vitest'; import { parseChainOutput, type TaskDefinition } from '../src/orchestrator/task-executor.js'; describe('parseChainOutput', () => { it('should parse a single task with <details> block', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend engineer with expertise in Node.js and TypeScript. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <details> <summary>Click to expand</summary> <prompt> # Task: Implement authentication ## MANDATORY RULES - Rule 1: Use TypeScript - Rule 2: Follow best practices ## OUTPUT Deliver working code. </prompt> </details> **Dependencies:** None **Estimated Duration:** 2-3 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); expect(tasks[0]).toMatchObject({ id: 'task-1.1', title: 'task-1.1', agentRoleDescription: 'Backend engineer with expertise in Node.js and TypeScript.', recommendedModel: 'GPT-4.1', dependencies: [], estimatedDuration: '2-3 hours', parallelGroup: undefined, }); expect(tasks[0].prompt).toContain('# Task: Implement authentication'); expect(tasks[0].prompt).toContain('Use TypeScript'); }); it('should parse a task without <details> block', () => { const markdown = ` **Task ID:** task-2.1 **Agent Role Description:** Frontend developer with React expertise. **Recommended Model:** Claude Sonnet 4 **Optimized Prompt:** <prompt> # Build Login Component Create a login form with email and password fields. </prompt> **Dependencies:** task-1.1 **Estimated Duration:** 1-2 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); expect(tasks[0]).toMatchObject({ id: 'task-2.1', agentRoleDescription: 'Frontend developer with React expertise.', recommendedModel: 'Claude Sonnet 4', dependencies: ['task-1.1'], estimatedDuration: '1-2 hours', }); }); it('should parse multiple tasks', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend architect. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Design the system. </prompt> **Dependencies:** None **Estimated Duration:** 3 hours --- **Task ID:** task-1.2 **Agent Role Description:** Database engineer. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Create schema. </prompt> **Dependencies:** task-1.1 **Estimated Duration:** 2 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(2); expect(tasks[0].id).toBe('task-1.1'); expect(tasks[1].id).toBe('task-1.2'); expect(tasks[1].dependencies).toEqual(['task-1.1']); }); it('should parse tasks with multiple dependencies', () => { const markdown = ` **Task ID:** task-3.1 **Agent Role Description:** Integration specialist. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Integrate components. </prompt> **Dependencies:** task-1.1, task-2.1, task-2.2 **Estimated Duration:** 4 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); expect(tasks[0].dependencies).toEqual(['task-1.1', 'task-2.1', 'task-2.2']); }); it('should parse tasks with parallel groups', () => { const markdown = ` **Task ID:** task-1.1 **Parallel Group:** 1 **Agent Role Description:** Backend developer A. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Implement feature A. </prompt> **Dependencies:** None **Estimated Duration:** 2 hours --- **Task ID:** task-1.2 **Parallel Group:** 1 **Agent Role Description:** Backend developer B. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Implement feature B. </prompt> **Dependencies:** None **Estimated Duration:** 2 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(2); expect(tasks[0].parallelGroup).toBe(1); expect(tasks[1].parallelGroup).toBe(1); }); it('should handle "none" dependencies case-insensitively', () => { const markdownLower = ` **Task ID:** task-1.1 **Agent Role Description:** Developer. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Task </prompt> **Dependencies:** none **Estimated Duration:** 1 hour `; const markdownUpper = markdownLower.replace('none', 'None'); const tasksLower = parseChainOutput(markdownLower); const tasksUpper = parseChainOutput(markdownUpper); expect(tasksLower[0].dependencies).toEqual([]); expect(tasksUpper[0].dependencies).toEqual([]); }); it('should return empty array for invalid markdown', () => { const markdown = ` # Some Random Markdown This doesn't match the expected format at all. `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(0); }); it('should handle tasks with complex multi-line prompts', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Full-stack developer. **Recommended Model:** Claude Sonnet 4 **Optimized Prompt:** <details> <summary>Click to expand</summary> <prompt> # Complex Task ## MANDATORY RULES **RULE #0: VERIFY CONTEXT ASSUMPTIONS FIRST** Before starting, verify these: - ✅ Check \`README.md\` - ✅ Check \`package.json\` **RULE #1: IMPLEMENTATION** - Step 1: Do this - Step 2: Do that ## OUTPUT FORMAT 1. Code 2. Tests 3. Documentation ## VERIFICATION CHECKLIST - [ ] Tests pass - [ ] Code is documented </prompt> </details> **Dependencies:** None **Estimated Duration:** 4-6 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); expect(tasks[0].prompt).toContain('RULE #0'); expect(tasks[0].prompt).toContain('VERIFICATION CHECKLIST'); expect(tasks[0].prompt).toContain('Tests pass'); }); it('should trim whitespace from all fields', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend developer with lots of spaces. **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Task with spaces </prompt> **Dependencies:** task-0.1 , task-0.2 **Estimated Duration:** 2 hours `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); expect(tasks[0].id).toBe('task-1.1'); expect(tasks[0].agentRoleDescription).toBe('Backend developer with lots of spaces.'); expect(tasks[0].recommendedModel).toBe('GPT-4.1'); expect(tasks[0].dependencies).toEqual(['task-0.1', 'task-0.2']); expect(tasks[0].estimatedDuration).toBe('2 hours'); }); it('should extract QC Agent Role', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend developer with Node.js expertise **Recommended Model:** Claude Sonnet 4 **Optimized Prompt:** <prompt> Build a REST API </prompt> **Dependencies:** None **Estimated Duration:** 2 hours **QC Agent Role:** Senior QA engineer with API testing expertise and security auditing experience **Verification Criteria:** Security: - [ ] No hardcoded credentials **Max Retries:** 2 `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); const task = tasks[0]; expect(task.qcRole).toBeDefined(); expect(task.qcRole).toContain('Senior QA engineer'); expect(task.qcRole).toContain('security auditing'); }); it('should extract Verification Criteria with multiple sections', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend developer **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Build API </prompt> **Dependencies:** None **Estimated Duration:** 1 hour **QC Agent Role:** QA Engineer **Verification Criteria:** Security: - [ ] No hardcoded credentials - [ ] All endpoints authenticated Functionality: - [ ] All CRUD operations implemented **Max Retries:** 3 `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); const task = tasks[0]; expect(task.verificationCriteria).toBeDefined(); expect(task.verificationCriteria).toContain('Security:'); expect(task.verificationCriteria).toContain('No hardcoded credentials'); expect(task.verificationCriteria).toContain('Functionality:'); }); it('should extract Max Retries as number', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend developer **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Build API </prompt> **Dependencies:** None **Estimated Duration:** 1 hour **QC Agent Role:** QA Engineer **Verification Criteria:** - [ ] Test 1 **Max Retries:** 3 `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); const task = tasks[0]; expect(task.maxRetries).toBe(3); expect(typeof task.maxRetries).toBe('number'); }); it('should default maxRetries to 2 when not specified', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend developer **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Build API </prompt> **Dependencies:** None **Estimated Duration:** 1 hour `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); const task = tasks[0]; expect(task.maxRetries).toBe(2); }); it('should handle tasks without QC fields (legacy format)', () => { const markdown = ` **Task ID:** task-1.1 **Agent Role Description:** Backend developer **Recommended Model:** GPT-4.1 **Optimized Prompt:** <prompt> Build API </prompt> **Dependencies:** None **Estimated Duration:** 1 hour `; const tasks = parseChainOutput(markdown); expect(tasks).toHaveLength(1); const task = tasks[0]; expect(task.qcRole).toBeUndefined(); expect(task.verificationCriteria).toBeUndefined(); expect(task.maxRetries).toBe(2); // Still has default }); });