sapper-iq 1.0.8 → 1.0.10

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sapper-iq",
-  "version": "1.0.8",
+  "version": "1.0.10",
   "description": "AI-powered development assistant that executes commands and builds projects",
   "main": "sapper.mjs",
   "bin": {

package/sapper.mjs

@@ -10,8 +10,13 @@ import { dirname, join } from 'path';
 
 const __filename = fileURLToPath(import.meta.url);
 const __dirname = dirname(__filename);
-const packageJson = JSON.parse(fs.readFileSync(join(__dirname, 'package.json'), 'utf8'));
-const CURRENT_VERSION = packageJson.version;
+
+// Initialize versioning
+let CURRENT_VERSION = "1.1.0";
+try {
+  const pkg = JSON.parse(fs.readFileSync(join(__dirname, 'package.json'), 'utf8'));
+  CURRENT_VERSION = pkg.version;
+} catch (e) {}
 
 const spinner = ora();
 const CONTEXT_FILE = '.sapper_context.json';
@@ -24,65 +29,13 @@ let rl = readline.createInterface({
   historySize: 100
 });
 
-// Helper function to safely prompt for input
 async function safeQuestion(query) {
   return new Promise((resolve) => {
     process.stdout.write(query);
-    rl.once('line', (answer) => {
-      resolve(answer.trim());
-    });
+    rl.once('line', (answer) => { resolve(answer.trim()); });
   });
 }
 
-// Helper function to check for updates
-async function checkForUpdates() {
-  try {
-    const response = await fetch('https://registry.npmjs.org/sapper-iq/latest');
-    const data = await response.json();
-    const latestVersion = data.version;
-    
-    if (latestVersion && latestVersion !== CURRENT_VERSION) {
-      console.log(chalk.yellow('🔄 UPDATE AVAILABLE!'));
-      console.log(chalk.gray(`   Current: v${CURRENT_VERSION}`));
-      console.log(chalk.green(`   Latest:  v${latestVersion}`));
-      console.log(chalk.cyan('   Run: npm update -g sapper-iq\n'));
-    }
-  } catch (error) {
-    // Silently fail if update check fails
-  }
-}
-
-// Helper function to update sapper
-async function updateSapper() {
-  console.log(chalk.cyan('🔄 Updating Sapper...'));
-  const confirm = await safeQuestion(chalk.yellow('Continue with update? (y/n): '));
-  if (confirm.toLowerCase() === 'y') {
-    return new Promise((resolve) => {
-      const proc = spawn('npm', ['update', '-g', 'sapper-iq'], { 
-        stdio: 'inherit' 
-      });
-      
-      proc.on('close', (code) => {
-        recreateReadline();
-        if (code === 0) {
-          console.log(chalk.green('\n✅ Sapper updated successfully!'));
-          console.log(chalk.gray('Please restart Sapper to use the new version.\n'));
-        } else {
-          console.log(chalk.red('\n❌ Update failed. Try manually: npm update -g sapper-iq\n'));
-        }
-        resolve();
-      });
-      
-      proc.on('error', (err) => {
-        recreateReadline();
-        console.log(chalk.red(`\n❌ Update error: ${err.message}\n`));
-        resolve();
-      });
-    });
-  }
-}
-
-// Helper function to recreate readline after shell commands
 function recreateReadline() {
   rl.close();
   rl = readline.createInterface({ 
@@ -93,659 +46,115 @@ function recreateReadline() {
   });
 }
 
-// Specialized prompt profiles for different tasks
-const PROMPT_PROFILES = {
-  documentation: `
-🎯 DOCUMENTATION MODE ACTIVATED
-Write comprehensive technical documentation. Execute tools to analyze codebase first.
-
-Documentation Types & Formats:
-
-**Use Cases:**
-Title: [Action User Wants to Perform]
-- Actor: [Who performs the action]
-- Preconditions: [What must be true before]
-- Main Flow:
-  1. User does X
-  2. System responds with Y
-  3. User confirms Z
-- Postconditions: [Result after completion]
-- Alternative Flows: [Error cases, edge cases]
-
-**User Stories:**
-"As a [role], I want [feature] so that [benefit]"
-- Acceptance Criteria:
-  - Given [context]
-  - When [action]
-  - Then [result]
-
-**BRD (Business Requirements Document):**
-## Background
-## Problem Statement
-## Objectives
-## Functional Requirements (numbered list)
-## Non-Functional Requirements (performance, security, scalability)
-## Success Metrics
-## Timeline & Milestones
-
-**Technical Documentation:**
-- API endpoints with request/response examples
-- Architecture diagrams (use mermaid)
-- Setup instructions
-- Environment variables
-
-Execute [TOOL:LIST] and [TOOL:READ] to understand project before documenting.`,
-
-  backend: `
-🎯 BACKEND MODE ACTIVATED - Node.js Specialist
-Build server-side APIs, business logic, and integrations using Node.js/Express.
-
-Tech Stack & Standards:
-- Framework: Express.js or Fastify
-- Runtime: Node.js (ES6+ with async/await)
-- Architecture: MVC, Clean Architecture, or layered approach
-
-Implementation Checklist:
-✓ RESTful API Design:
-  - GET /api/resource (list/read)
-  - POST /api/resource (create)
-  - PUT /PATCH /api/resource/:id (update)
-  - DELETE /api/resource/:id (delete)
-  - Return proper status codes: 200, 201, 400, 401, 404, 500
-
-✓ Request Handling:
-  - Use express.json() for body parsing
-  - Validate inputs with joi or zod
-  - Implement error middleware
-  - Add request logging (morgan/winston)
-
-✓ Security:
-  - Use helmet for security headers
-  - Implement CORS properly
-  - JWT authentication with bcrypt password hashing
-  - Rate limiting with express-rate-limit
-  - Input sanitization
-
-✓ Project Structure:
-/src
-  /routes     - API endpoints
-  /controllers - Request handlers
-  /services   - Business logic
-  /models     - Data models
-  /middleware - Auth, validation, errors
-  /config     - Environment configs
-  /utils      - Helper functions
-
-✓ Best Practices:
-  - Use environment variables (.env with dotenv)
-  - Implement graceful shutdown
-  - Add health check endpoint: GET /health
-  - Use async/await with try-catch
-  - Create separate router files
-
-Execute tools to create Express server structure immediately.`,
-
-  frontend: `
-🎯 FRONTEND MODE ACTIVATED
-Focus on UI components, styling, and user interactions.
-
-Implementation Rules:
-- Component-based architecture (React/Vue/Svelte)
-- Responsive design with mobile-first approach
-- Use Tailwind CSS or styled-components
-- Implement proper state management
-- Add loading states and error boundaries
-- Handle forms with validation
-- Optimize for performance (lazy loading, memoization)
-
-Execute tools to create component structure immediately.`,
-
-  testing: `
-🎯 TESTING MODE ACTIVATED
-Write comprehensive test suites with high coverage.
-
-Test Requirements:
-- Unit tests: Test individual functions/components
-- Integration tests: Test feature workflows
-- Use describe/it blocks with clear names
-- Add assertions: expect().toBe(), toEqual(), toHaveBeenCalled()
-- Mock external dependencies
-- Test edge cases and error scenarios
-- Aim for 80%+ coverage
-
-Execute tools to create test files immediately.`,
-
-  database: `
-🎯 DATABASE MODE ACTIVATED - PostgreSQL Specialist
-Design schemas, write migrations, and optimize queries for PostgreSQL.
-
-Tech Stack:
-- Database: PostgreSQL 14+
-- ORM Options: Prisma (recommended), Sequelize, or TypeORM
-- Query Builder: Knex.js
-- Migrations: Prisma Migrate or Knex migrations
-
-Schema Design Best Practices:
-✓ Data Types:
-  - Use SERIAL or UUID for primary keys
-  - TEXT for variable strings (not VARCHAR)
-  - TIMESTAMP WITH TIME ZONE for dates
-  - JSONB for flexible data (indexed)
-  - ENUM types for fixed choices
-
-✓ Table Structure:
-CREATE TABLE users (
-  id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
-  email TEXT UNIQUE NOT NULL,
-  password_hash TEXT NOT NULL,
-  created_at TIMESTAMP WITH TIME ZONE DEFAULT NOW(),
-  updated_at TIMESTAMP WITH TIME ZONE DEFAULT NOW()
-);
-
-✓ Relationships:
-  - Foreign Keys: REFERENCES other_table(id) ON DELETE CASCADE
-  - Indexes on FK columns: CREATE INDEX idx_user_id ON posts(user_id)
-  - Join tables for many-to-many
-
-✓ Performance:
-  - Add indexes on frequently queried columns
-  - Use partial indexes for conditional queries
-  - Create composite indexes for multi-column queries
-  - Analyze query plans: EXPLAIN ANALYZE
-
-✓ Migrations Pattern:
-  - Up: Create/alter tables and indexes
-  - Down: Rollback changes safely
-  - Version control all migrations
-  - Never edit existing migrations
-
-✓ Prisma Schema Example:
-model User {
-  id        String   @id @default(uuid())
-  email     String   @unique
-  posts     Post[]
-  createdAt DateTime @default(now())
-  updatedAt DateTime @updatedAt
-}
-
-✓ Connection:
-  - Use connection pooling (pg.Pool)
-  - Set max connections: 20-50
-  - Use prepared statements
-  - Handle connection errors gracefully
-
-✓ Seed Data:
-  - Create seed files for development
-  - Use transactions for bulk inserts
-  - Make seeds idempotent
-
-Execute tools to create schema files and migrations immediately.`,
-
-  devops: `
-🎯 DEVOPS MODE ACTIVATED
-Setup CI/CD, containerization, deployment automation.
-
-DevOps Tasks:
-- Write Dockerfile with multi-stage builds
-- Create docker-compose.yml for local dev
-- Setup GitHub Actions or GitLab CI
-- Add environment-specific configs
-- Implement health checks and monitoring
-- Use secrets management
-- Optimize build times
-
-Execute tools to create config files immediately.`
-};
-
-// Auto-detect task profile from user input
-function detectTaskProfile(input) {
-  const text = input.toLowerCase();
-  
-  // Documentation keywords
-  if (text.match(/\b(document|documentation|readme|brd|prd|user stor(y|ies)|use case|spec|wiki|guide)\b/))
-    return 'documentation';
-  
-  // Backend keywords
-  if (text.match(/\b(api|endpoint|backend|server|route|controller|middleware|rest|graphql|service)\b/))
-    return 'backend';
-  
-  // Frontend keywords
-  if (text.match(/\b(ui|frontend|component|page|design|style|css|tailwind|react|vue|svelte|button|form|navbar)\b/))
-    return 'frontend';
-  
-  // Testing keywords
-  if (text.match(/\b(test|testing|spec|jest|vitest|cypress|unit test|integration|e2e|coverage)\b/))
-    return 'testing';
-  
-  // Database keywords
-  if (text.match(/\b(database|schema|migration|model|orm|sql|postgres|mysql|mongo|prisma|sequelize)\b/))
-    return 'database';
-  
-  // DevOps keywords
-  if (text.match(/\b(docker|container|ci\/cd|deploy|pipeline|kubernetes|k8s|helm|terraform|ansible)\b/))
-    return 'devops';
-  
-  return null;
-}
-
-// --- Tool Logic ---
 const tools = {
   read: (path) => {
-    try {
-      return fs.readFileSync(path.trim(), 'utf8');
-    } catch (error) {
-      return `Error reading file: ${error.message}`;
-    }
+    try { return fs.readFileSync(path.trim(), 'utf8'); } 
+    catch (error) { return `Error reading file: ${error.message}`; }
   },
   write: (path, content) => {
     try {
       fs.writeFileSync(path.trim(), content);
       return `Successfully saved changes to ${path}`;
-    } catch (error) {
-      return `Error writing file: ${error.message}`;
-    }
+    } catch (error) { return `Error writing file: ${error.message}`; }
   },
   mkdir: (path) => {
     try {
       fs.mkdirSync(path.trim(), { recursive: true });
       return `Directory created: ${path}`;
-    } catch (error) {
-      return `Error creating directory: ${error.message}`;
-    }
+    } catch (error) { return `Error creating directory: ${error.message}`; }
   },
   shell: async (cmd) => {
     console.log(chalk.red.bold(`\n[SECURITY] Sapper wants to execute: `) + chalk.white(cmd));
     const confirm = await safeQuestion(chalk.yellow('Allow? (y/n): '));
     if (confirm.toLowerCase() === 'y') {
       return new Promise((resolve) => {
-        // Use shell for complex commands with pipes, redirects, cd, &&, ||, etc
-        const useShell = cmd.includes('&&') || cmd.includes('||') || cmd.includes('|') || cmd.includes('cd ') || cmd.includes('>');
-        
+        const useShell = cmd.includes('&&') || cmd.includes('|') || cmd.includes('cd ');
         console.log(chalk.cyan(`\n[RUNNING] ${cmd}\n`));
-        
-        let proc;
-        if (useShell) {
-          // For complex commands, use shell
-          proc = spawn('sh', ['-c', cmd], { 
-            stdio: 'inherit',
-            shell: true 
-          });
-        } else {
-          // For simple commands, parse and use direct execution
-          const parts = cmd.trim().split(/\s+/);
-          const executable = parts[0];
-          const args = parts.slice(1);
-          proc = spawn(executable, args, { 
-            stdio: 'inherit',
-            shell: false 
-          });
-        }
-        
-        proc.on('close', (code) => {
-          // Recreate readline after shell command completes
-          recreateReadline();
-          console.log(chalk.green(`\n[✓] Command completed with exit code ${code}\n`));
-          resolve(`Command completed with exit code ${code}.`);
+        const proc = spawn(useShell ? 'sh' : cmd.split(' ')[0], useShell ? ['-c', cmd] : cmd.split(' ').slice(1), { 
+          stdio: 'inherit', shell: useShell 
         });
-        
-        proc.on('error', (err) => {
+        proc.on('close', (code) => {
           recreateReadline();
-          console.log(chalk.red(`\n[✗] Command error: ${err.message}\n`));
-          resolve(`Execution Error: ${err.message}`);
+          resolve(`Command completed with code ${code}`);
         });
       });
     }
     return "Command blocked by user.";
   },
   list: (path) => {
-    try {
-      return fs.readdirSync(path || '.').join('\n');
-    } catch (error) {
-      return `Error listing directory: ${error.message}`;
-    }
-  },
-  search: (pattern) => {
-    try {
-      const { execSync } = require('child_process');
-      const cmd = `grep -rnEi "${pattern.trim()}" . --exclude-dir=node_modules --exclude-dir=.git`;
-      return execSync(cmd, { encoding: 'utf8' }) || "No matches found.";
-    } catch (e) { 
-      return "No matches found."; 
-    }
+    try { return fs.readdirSync(path.trim() || '.').join('\n'); } 
+    catch (e) { return `Error: ${e.message}`; }
   }
 };
 
-async function selectModel() {
-  try {
-    const localModels = await ollama.list();
-    if (localModels.models.length === 0) {
-      console.log(chalk.red('❌ No Ollama models found!'));
-      console.log(chalk.yellow('Please install at least one model:'));
-      console.log(chalk.gray('  ollama pull llama2'));
-      console.log(chalk.gray('  ollama pull codellama'));
-      process.exit(1);
-    }
-    console.log(chalk.magenta.bold("\nAvailable Models:"));
-    localModels.models.forEach((m, i) => console.log(`${i + 1}. ${chalk.white(m.name)}`));
-    const choice = await safeQuestion(chalk.yellow('\nChoose model: '));
-    const index = parseInt(choice) - 1;
-    return localModels.models[index]?.name || localModels.models[0].name;
-  } catch (error) {
-    console.log(chalk.red('❌ Failed to connect to Ollama!'));
-    console.log(chalk.yellow('Please make sure Ollama is running:'));
-    console.log(chalk.gray('  1. Install Ollama: https://ollama.ai'));
-    console.log(chalk.gray('  2. Start Ollama: ollama serve'));
-    console.log(chalk.gray('  3. Install a model: ollama pull llama2'));
-    console.log(chalk.red(`\nError details: ${error.message}`));
-    process.exit(1);
-  }
-}
-
 async function runSapper() {
   console.clear();
-  console.log(chalk.cyan.bold(` SAPPER v${CURRENT_VERSION} | Multi-Tool Execution Mode`));
-  console.log(chalk.gray("Commands: /reset, /session-info, /step, /version, /update, /help, exit\n"));
-
-  // Check for updates on startup
-  await checkForUpdates();
-
-  // Early Ollama connectivity check
-  console.log(chalk.gray('🔍 Checking Ollama connection...'));
+  console.log(chalk.cyan.bold(` SAPPER v${CURRENT_VERSION} | Autonomous "OpenCode" Mode`));
   
   let messages = [];
   if (fs.existsSync(CONTEXT_FILE)) {
     const resume = await safeQuestion(chalk.green('Resume previous session? (y/n): '));
-    if (resume.toLowerCase() === 'y') {
-      messages = JSON.parse(fs.readFileSync(CONTEXT_FILE, 'utf8'));
-    }
+    if (resume.toLowerCase() === 'y') messages = JSON.parse(fs.readFileSync(CONTEXT_FILE, 'utf8'));
   }
 
-  const selectedModel = await selectModel();
+  const localModels = await ollama.list();
+  localModels.models.forEach((m, i) => console.log(`${i + 1}. ${m.name}`));
+  const choice = await safeQuestion(chalk.yellow('\nChoose model: '));
+  const selectedModel = localModels.models[parseInt(choice) - 1]?.name || localModels.models[0].name;
 
   if (messages.length === 0) {
     messages = [{
       role: 'system',
-      content: `You are Sapper, a senior software engineer AI assistant. Execute tools to COMPLETE tasks fully.
-
-**COMPLETE TASK WORKFLOW:**
-When user says "analyze files" → DO ALL THESE STEPS IN ONE RESPONSE:
-1. [TOOL:LIST]./[/TOOL] - see what files exist
-2. [TOOL:READ]./file1.md[/TOOL] - read each relevant file
-3. [TOOL:READ]./file2.md[/TOOL] - read more files as needed
-4. Provide detailed analysis based on what you read
-5. [SUMMARY:Analyzed X files, found Y patterns, created Z documentation]
-
-❌ WRONG (incomplete):
-[TOOL:LIST]./[/TOOL]
-(stops here - no reading, no analysis)
-
-✅ CORRECT (complete):
-[TOOL:LIST]./[/TOOL]
-[TOOL:READ]./README.md[/TOOL]
-[TOOL:READ]./docs.md[/TOOL]
-
-Based on the files:
-- README.md contains project setup instructions
-- docs.md has API documentation for 5 endpoints
-- The project uses Express.js with PostgreSQL
-
-[SUMMARY:Analyzed 2 documentation files covering setup and API endpoints]
-
-**TOOL FORMAT:**
-[TOOL:TYPE]path]content[/TOOL]
-
-**Available Tools:**
-- SHELL: Execute commands
-- READ: Read file contents (use this OFTEN!)
-- WRITE: Create/update files
-- MKDIR: Create directories
-- LIST: List directory contents
-- SEARCH: Search for patterns
-
-**Format Examples:**
-[TOOL:SHELL]npm install[/TOOL]
-[TOOL:READ]./package.json[/TOOL]
-[TOOL:WRITE]./app.js]console.log('hello')[/TOOL]
-[TOOL:LIST]./[/TOOL]
-[TOOL:SEARCH]function auth[/TOOL]
-
-**Multi-line files:**
-[TOOL:WRITE]./README.md]
-# Title
-- [ ] checkbox
-Arrays: [1, 2, 3]
-[/TOOL]
-
-**Multiple Tools Per Response:**
-You MUST execute ALL necessary tools in ONE response. Example:
-[TOOL:MKDIR]./src[/TOOL]
-[TOOL:WRITE]./src/server.js]const express = require('express')[/TOOL]
-[TOOL:WRITE]./package.json]{"name": "app"}[/TOOL]
-
-Created project structure with server and package.json.
-[SUMMARY:Created Node.js project with Express server]
-
-**Shell Commands:**
-[TOOL:SHELL]cd /path && npm install && npm start[/TOOL]
-
-**Critical Rules:**
-1. Execute ALL needed tools in your response
-2. Read files after listing them
-3. Provide analysis/explanation after reading
-4. End with [SUMMARY:what you completed]
-5. NEVER just execute one tool and stop
-
-Working directory: ${process.cwd()}`
+      content: `You are Sapper, a senior engineer. 
+      STRATEGY:
+      1. When asked to analyze, use [TOOL:LIST]./[/TOOL] first.
+      2. Immediately [TOOL:READ] key files (package.json, README.md, entry points) in the SAME turn.
+      3. Use the format: [TOOL:TYPE]path]content[/TOOL]. For LIST/READ, content is empty.
+      4. DO NOT ask for permission to read or list. Just do it.`
     }];
   }
 
-  // Display working directory awareness
-  console.log(chalk.yellow(`Working Directory: ${process.cwd()}\n`));
-
   const ask = () => {
     safeQuestion(chalk.blue.bold('\nIbrahim ➔ ')).then(async (input) => {
       if (input.toLowerCase() === 'exit') process.exit();
-      if (input.toLowerCase() === '/reset' || input.toLowerCase() === '/clear-session') {
-        if (fs.existsSync(CONTEXT_FILE)) {
-          const fileSize = fs.statSync(CONTEXT_FILE).size;
-          console.log(chalk.yellow(`\n🗑️  Clearing session (${(fileSize / 1024).toFixed(2)}KB)...`));
-          fs.unlinkSync(CONTEXT_FILE);
-          console.log(chalk.green('✅ Session cleared! Starting fresh...\n'));
-        } else {
-          console.log(chalk.yellow('\nℹ️  No session to clear.\n'));
-        }
-        return runSapper();
-      }
-      if (input.toLowerCase() === '/session-info') {
-        if (fs.existsSync(CONTEXT_FILE)) {
-          const data = JSON.parse(fs.readFileSync(CONTEXT_FILE, 'utf8'));
-          const fileSize = fs.statSync(CONTEXT_FILE).size;
-          console.log(chalk.cyan(`\n📊 Session Info:`));
-          console.log(chalk.gray(`  Messages: ${data.length}`));
-          console.log(chalk.gray(`  File Size: ${(fileSize / 1024).toFixed(2)}KB`));
-          console.log(chalk.gray(`  Last Message: ${data[data.length - 1]?.role || 'N/A'}`));
-        } else {
-          console.log(chalk.yellow('\nℹ️  No active session.\n'));
-        }
-        return ask();
-      }
-      if (input.toLowerCase() === '/version') {
-        console.log(chalk.cyan(`\n📦 Sapper Version: v${CURRENT_VERSION}`));
-        console.log(chalk.gray(`   Node.js: ${process.version}`));
-        console.log(chalk.gray(`   Platform: ${process.platform}\n`));
-        // Check for updates
-        await checkForUpdates();
-        return ask();
-      }
-      if (input.toLowerCase() === '/update') {
-        await updateSapper();
-        return ask();
-      }
-      if (input.toLowerCase() === '/step') {
-        stepMode = !stepMode;
-        console.log(chalk.yellow(`Step Mode is ${stepMode ? 'ON' : 'OFF'}`));
-        return ask();
-      }
-      if (input.toLowerCase() === '/help') {
-        console.log(chalk.cyan(`\n📚 Sapper Commands:`));
-        console.log(chalk.gray(`  /reset or /clear-session  - Start a new session`));
-        console.log(chalk.gray(`  /session-info            - Show current session details`));
-        console.log(chalk.gray(`  /version                 - Show version and check for updates`));
-        console.log(chalk.gray(`  /update                  - Update Sapper to latest version`));
-        console.log(chalk.gray(`  /step                    - Toggle step-by-step mode`));
-        console.log(chalk.gray(`  /help                    - Show this help menu`));
-        console.log(chalk.gray(`  exit                     - Exit Sapper\n`));
-        return ask();
-      }
-
-      // Auto-detect task profile and inject specialized instructions
-      const profile = detectTaskProfile(input);
-      let contextMsg = input;
-      
-      // Inject profile if detected
-      if (profile) {
-        console.log(chalk.magenta(`\n🎯 ${profile.toUpperCase()} MODE DETECTED\n`));
-        contextMsg = `${input}\n\n${PROMPT_PROFILES[profile]}`;
-      }
-      
-      // Check if user mentioned a directory and provide context
-      const dirMatch = input.match(/\/Users\/[^\s]+|\/[a-zA-Z0-9_\/-]+/g);
-      
-      if (dirMatch && dirMatch[0]) {
-        const mentionedDir = dirMatch[0];
-        try {
-          if (fs.existsSync(mentionedDir) && fs.statSync(mentionedDir).isDirectory()) {
-            const files = fs.readdirSync(mentionedDir).slice(0, 10).join(', ');
-            contextMsg += `\n\n[CONTEXT: Directory "${mentionedDir}" contains: ${files}${fs.readdirSync(mentionedDir).length > 10 ? '...' : ''}]`;
-          }
-        } catch (e) {
-          // Silently ignore if directory doesn't exist
-        }
-      }
-
-      messages.push({ role: 'user', content: contextMsg });
+      messages.push({ role: 'user', content: input });
 
       let active = true;
-      let iterations = 0;
-      while (active && iterations < 30) {
-        iterations++;
-
-        if (stepMode) {
-          const proceed = await safeQuestion(chalk.gray('\n[STEP-MODE] Press Enter to continue (or type "/stop"): '));
-          if (proceed.toLowerCase() === '/stop') break;
-        }
-
-        spinner.stop();
-        console.log(chalk.blue(`\n${selectedModel} is thinking...`));
-        
-        let response;
-        try {
-          response = await ollama.chat({ 
-            model: selectedModel, 
-            messages, 
-            stream: true,
-            options: { num_ctx: 16384 } 
-          });
-        } catch (error) {
-          console.log(chalk.red('\n❌ Failed to communicate with Ollama!'));
-          console.log(chalk.yellow('Possible issues:'));
-          console.log(chalk.gray('  - Ollama service stopped'));
-          console.log(chalk.gray('  - Model was removed'));
-          console.log(chalk.gray('  - Network connection issue'));
-          console.log(chalk.red(`Error: ${error.message}`));
-          console.log(chalk.cyan('\n💡 Try restarting Sapper or check Ollama status'));
-          active = false;
-          ask();
-          return;
-        }
+      while (active) {
+        if (stepMode) await safeQuestion(chalk.gray('[STEP] Press Enter to let AI think...'));
         
+        spinner.start('Thinking...');
+        const response = await ollama.chat({ model: selectedModel, messages, stream: true });
+        spinner.stop();
+
         let msg = '';
         process.stdout.write(chalk.white('Sapper: '));
-        
-        try {
-          for await (const chunk of response) {
-            if (chunk.message && chunk.message.content) {
-              process.stdout.write(chunk.message.content);
-              msg += chunk.message.content;
-            }
-          }
-        } catch (error) {
-          console.log(chalk.red('\n\n❌ Connection interrupted while streaming response!'));
-          console.log(chalk.yellow(`Error: ${error.message}`));
-          console.log(chalk.cyan('💡 The conversation will continue, but you may want to restart Sapper'));
-          msg += `\n[ERROR: Response interrupted - ${error.message}]`;
+        for await (const chunk of response) {
+          process.stdout.write(chunk.message.content);
+          msg += chunk.message.content;
         }
         console.log();
-        
         messages.push({ role: 'assistant', content: msg });
 
-        const summaryMatch = msg.match(/\[SUMMARY:(.*?)\]/s);
+        const toolMatches = [...msg.matchAll(/\[TOOL:(\w+)\]([^\]\n]+)(?:\]([\s\S]*?))?\[\/TOOL\]/g)];
         
-        // Primary format: [TOOL:TYPE:path]content[/TOOL]
-        const toolMatches = [...msg.matchAll(/\[TOOL:(\w+)\]([^\[]+?)\]([\s\S]*?)\[\/TOOL\]/g)];
-
-        if (summaryMatch) {
-          console.log(chalk.green.bold("\n✅ MISSION COMPLETE:"));
-          console.log(chalk.white(summaryMatch[1].trim()));
-          active = false;
-          continue;
-        }
-
         if (toolMatches.length > 0) {
           for (const match of toolMatches) {
-            const [_, name, path, content] = match;
-            const toolName = name.toLowerCase();
-            console.log(chalk.cyan(`\n[ACTION] Executing ${toolName} on: ${path}`));
+            const [_, type, path, content] = match;
+            console.log(chalk.cyan(`\n[ACTION] ${type} -> ${path}`));
             
             let result;
-            try {
-              if (toolName === 'shell') result = await tools.shell(path);
-              else if (toolName === 'write') result = tools.write(path, content);
-              else if (toolName === 'mkdir') result = tools.mkdir(path);
-              else if (toolName === 'read') result = tools.read(path);
-              else if (toolName === 'list') result = tools.list(path);
-              else if (toolName === 'search') result = tools.search(path);
-              else result = `Unknown tool: ${name}`;
-            } catch (e) {
-              result = `Error: ${e.message}`;
-            }
-            
-            console.log(chalk.gray(`> Result: ${result.substring(0, 60)}...`));
-            messages.push({ role: 'user', content: `TOOL_RESULT for ${path}: ${result}` });
+            if (type.toLowerCase() === 'list') result = tools.list(path);
+            else if (type.toLowerCase() === 'read') result = tools.read(path);
+            else if (type.toLowerCase() === 'mkdir') result = tools.mkdir(path);
+            else if (type.toLowerCase() === 'write') result = tools.write(path, content);
+            else if (type.toLowerCase() === 'shell') result = await tools.shell(path);
+
+            messages.push({ role: 'user', content: `RESULT (${path}): ${result}` });
           }
           fs.writeFileSync(CONTEXT_FILE, JSON.stringify(messages));
-          
-          // Add interrupt check after tool execution
-          console.log(chalk.gray('\n[Press Enter to continue or type "/stop" to halt execution]'));
-          const userChoice = await safeQuestion('');
-          if (userChoice.toLowerCase() === '/stop') {
-            console.log(chalk.yellow('\n⏹️  Execution halted by user'));
-            active = false;
-            break;
-          }
         } else {
-          const planMatch = msg.match(/\[PLAN:([\s\S]*?)\]/) || msg.match(/\[PLAN\]([\s\S]*?)\[\/PLAN\]/);
-          if (planMatch) {
-            const feedback = await safeQuestion(chalk.yellow('\nModify plan or type "go": '));
-            if (feedback.toLowerCase() === '/stop') { active = false; break; }
-            messages.push({ role: 'user', content: feedback.toLowerCase() === 'go' ? "Plan approved. Proceed with all steps." : feedback });
-          } else {
-            active = false;
-          }
-        }
-
-        // Safety check: if model is repeating itself, break the loop
-        if (iterations > 5) {
-          const recentMessages = messages.slice(-4);
-          const isRepeating = recentMessages.every(m => 
-            m.role === 'assistant' && 
-            recentMessages[0].content && 
-            m.content === recentMessages[0].content
-          );
-          if (isRepeating) {
-            console.log(chalk.yellow('\n⚠️  Detected repetitive behavior, stopping execution'));
-            active = false;
-          }
+          active = false;
         }
       }
       ask();
@@ -754,4 +163,4 @@ Working directory: ${process.cwd()}`
   ask();
 }
 
-runSapper();
+runSapper();