@pageai/ralph-loop 1.1.0 → 1.4.0

package/.agent/PROMPT.md

@@ -4,7 +4,7 @@ You are implementing the project described in @.agent/prd/SUMMARY.md
 
 ## Required Setup
 
-Run `npm run dev` (as background process).
+Run `npm run dev` (as background process) in `src` directory.
 App will be running at http://localhost:3000
 
 ## Before Starting
@@ -29,7 +29,7 @@ Tasks are listed in @.agent/tasks.json
 8. All tests must pass. Broke unrelated test? Fix it before proceeding.
 9. When tests pass, set `passes: true` in `tasks.json` for the task you completed.
 10. Log entry → `.agent/logs/LOG.md` (date, brief summary, screenshot path)
-11. Update `.agent/STRUCTURE.md` if dirs changed
+11. Update `.agent/STRUCTURE.md` if dirs changed. Exclude dotfiles, tests and config.
 12. Commit changes, using the Conventional Commit format.
 
 ## Rules
@@ -42,15 +42,15 @@ Tasks are listed in @.agent/tasks.json
 ## Help Tags
 
 Try solving tasks yourself first.
-Only use when truly stuck.
+When stuck after all possible solutions exhausted, output one of the following tags:
 
-**BLOCKED** — technical issues: Playwright broken, deps won't install, env issues, no network, service outages, invalid credentials. Output:
+1. **BLOCKED** — technical issues: Playwright broken, dev server not working, deps won't install, env issues, no network, service outages, invalid/missing credentials. Output:
 
 ```
 <promise>BLOCKED:brief description</promise>
 ```
 
-**DECIDE** — need human input: lib choices, architecture, unclear requirements, breaking changes. Output:
+2. **DECIDE** — need human input: lib choices, architecture, unclear requirements, breaking changes. Output:
 
 ```
 <promise>DECIDE:question (Option A vs B)</promise>

package/.agent/STEERING.md

@@ -1,3 +1,5 @@
 # **Critical Steering Work**
 
-No critical work pending. All issues resolved.
+Go to the source code directory and install dependencies.
+
+Check that Playwright is running by starting the dev server and taking a screenshot. Save the screenshot to the .agent/screenshots directory.

package/.agent/skills/prd-creator/JSON.md

@@ -546,7 +546,7 @@ Acceptance criteria must be verifiable, not vague. "Works correctly" is bad. "Bu
 For any story with UI changes: Always include "Verify in browser using playwright" as acceptance criteria. This ensures visual verification of frontend work.
 
 **Important:**
-Tasks cannot be end to end tests or unit tests. These should be individual steps (verification criteria) in a task, if needed.
+End to end tests or unit tests are individual steps (verification criteria) in a task, but *CANNOT* be tasks themselves.
 
 ### Initialize All Tasks to `"passes": false`
 
@@ -580,6 +580,12 @@ Save the complete task list as `PROJECT_ROOT/.agent/tasks.json`:
 ]
 ```
 
+**Important:**
+When outputting tasks, create them, sequentially. Output them one by one so the user can review each.
+*DO NOT* output in parallel.
+*DO NOT* use background agents.
+*DO NOT* use subagents.
+
 ## After Generation
 
 Present the tasks to the user:

package/.agent/skills/prd-creator/PRD.md

@@ -10,6 +10,8 @@ Help beginner-level developers transform software ideas into comprehensive PRD.m
 4. Keep tone friendly and supportive, use plain language
 5. Track assumptions throughout - they'll go in the PRD
 
+**Important**: Grill the user for information to make sure you have all the required details in order to implement the PRD.
+
 ## Topics to Cover
 
 Gather information on these aspects:

package/.claude/settings.json

@@ -1,54 +1,43 @@
 {
-    "permissions": {
-        "deny": [
-            "Read(*.env)",
-            "Bash(sudo *)",
-            "Read(~/.ssh/**)",
-            "Read(.agent/history/**)"
-        ],
-        "allow": [
-            "**"
+  "permissions": {
+    "deny": ["Read(.agent/history/**)"],
+    "allow": ["**"]
+  },
+  "defaultMode": "acceptEdits",
+  "enableAllProjectMcpServers": true,
+  "enabledMcpjsonServers": ["playwright", "context7", "sequential-thinking"],
+  "hooks": {
+    "UserPromptSubmit": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node $CLAUDE_PROJECT_DIR/.claude/hooks/pre-tool-use.js"
+          }
         ]
-    },
-    "defaultMode": "acceptEdits",
-    "enableAllProjectMcpServers": true,
-    "enabledMcpjsonServers": [
-        "playwright",
-        "context7",
-        "sequential-thinking"
+      }
     ],
-    "hooks": {
-        "UserPromptSubmit": [
-            {
-                "hooks": [
-                    {
-                        "type": "command",
-                        "command": "node $CLAUDE_PROJECT_DIR/.claude/hooks/pre-tool-use.js"
-                    }
-                ]
-            }
-        ],
-        "PreToolUse": [
-            {
-                "matcher": "Edit|Write|Bash",
-                "hooks": [
-                    {
-                        "type": "command",
-                        "command": "node $CLAUDE_PROJECT_DIR/.claude/hooks/pre-tool-use.js"
-                    }
-                ]
-            }
-        ],
-        "Notification": [
-            {
-                "matcher": "",
-                "hooks": [
-                    {
-                        "type": "command",
-                        "command": "node $CLAUDE_PROJECT_DIR/.claude/hooks/play-sound.js notification"
-                    }
-                ]
-            }
+    "PreToolUse": [
+      {
+        "matcher": "Edit|Write|Bash",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node $CLAUDE_PROJECT_DIR/.claude/hooks/pre-tool-use.js"
+          }
         ]
-    }
+      }
+    ],
+    "Notification": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "node $CLAUDE_PROJECT_DIR/.claude/hooks/play-sound.js notification"
+          }
+        ]
+      }
+    ]
+  }
 }

package/AGENTS.md

@@ -41,4 +41,4 @@ Verify written code by:
 - Checking for type errors
 - Checking for lint errors
 - Smoke testing and checking for runtime errors with Playwright
-- Takins screenshots and verifying the UI is as expected
+- Taking screenshots and verifying the UI is as expected

package/README.md

@@ -7,12 +7,13 @@ This is an implementation that actually works, containing a hackable script so y
 ![Ralph Wiggum Loop](https://github.com/user-attachments/assets/052d5290-7e83-4bfb-a6b5-6be761cbe890)
 
 - [Getting Started](#getting-started)
-  - [Step 1: Install Ralph](#step-1-install-ralph)
-  - [Step 2: Create a PRD + task list](#step-2-create-a-prd--task-list)
-  - [Step 3: Set up the agent inside Docker sandbox](#step-3-set-up-the-agent-inside-docker-sandbox)
-  - [Step 4: Run Ralph](#step-4-run-ralph)
-  - [(optional) Adjusting to your language/framework](#optional-adjusting-to-your-languageframework)
+  - [(Optional) Set up code base](#optional-set-up-code-base)
+  - [1️⃣ Step 1: Install Ralph](#1️⃣-step-1-install-ralph)
+  - [2️⃣ Step 2: Create a PRD + task list](#2️⃣-step-2-create-a-prd--task-list)
+  - [3️⃣ Step 3: Set up the agent inside Docker sandbox](#3️⃣-step-3-set-up-the-agent-inside-docker-sandbox)
+  - [4️⃣ Step 4: Run Ralph](#4️⃣-step-4-run-ralph)
 - [Run the loop](#run-the-loop)
+  - [(Optional) Adjusting to your language/framework](#optional-adjusting-to-your-languageframework)
 - [How It Works](#how-it-works)
 - [How Is This Different from Other Ralphs?](#how-is-this-different-from-other-ralphs)
 - [Steering the Agent](#steering-the-agent)
@@ -27,11 +28,27 @@ This is an implementation that actually works, containing a hackable script so y
 - [Reference](#reference)
   - [Playwright configuration](#playwright-configuration)
   - [Vitest configuration](#vitest-configuration)
+  - [Running with a different agentic CLI](#running-with-a-different-agentic-cli)
+  - [Starting from scratch](#starting-from-scratch)
 - [License](#license)
 
+---------------------------------
+
 ## Getting Started
 
-### Step 1: Install Ralph
+### (Optional) Set up code base
+
+I recommend using a CLI to bootstrap your project with the necessary tools and dependencies, e.g.:
+
+```bash
+npx create-vite@latest src --template react-ts
+# or
+npx create-next-app@latest src
+```
+
+> If you must start from a blank slate, which is not recommended, see [Starting from scratch](#starting-from-scratch).
+
+### 1️⃣ Step 1: Install Ralph
 
 Run this in your project's directory to install Ralph.
 
@@ -39,13 +56,17 @@ Run this in your project's directory to install Ralph.
 npx @pageai/ralph-loop
 ```
 
-### Step 2: Create a PRD + task list
+### 2️⃣ Step 2: Create a PRD + task list
 
 Use the `prd-creator` skill to generate a PRD from your requirements.<br/>
 Open up Claude Code and prompt it with **your requirements**. Like so:
 
 ```
-Use the prd-creator skill to help me create a PRD and task list for these requirements:
+Use the prd-creator skill to help me create a PRD and task list for the below requirements.
+
+An app is already set up with Next.js, Tailwind CSS and TypeScript.
+
+Requirements:
 
 - A SaaS product that helps users manage their finances.
 - Target audience: Small business owners and freelancers.
@@ -57,16 +78,26 @@ Use the prd-creator skill to help me create a PRD and task list for these requir
   - Connect to bank accounts and credit cards.
   - Connect to accounting software.
   - Connect to payment processors.
-- Next.js web app with Tailwind CSS and TypeScript.
 - Use the shadcn/ui library for components.
+- Integrate with Stripe for payments.
+- Use Supabase for database.
+- You can find env variables in the .env.example file: SUPABASE_URL, SUPABASE_PUBLISHABLE_KEY, STRIPE_SECRET_KEY, etc. are available in the runtime.
 
 // etc.
 ```
 
+Pro tips:
+- mention libraries and frameworks you want to use
+- mention env variables, e.g. for database, 3rd party API keys, etc. Store them in a .env file that you add to **.gitignore**
+- describe user flows and journeys
+- add relevant docs and UI references if applicable inside `/docs` and mention them in the requirements
+- be as descriptive as possible
+- *it's fine to write in your own language*
+
 Then follow the Skill's instructions and verify the PRD and then tasks.<br/>
 **It is highly recommended that you review individual task requirements before starting the loop. Review EACH TASK INDIVIDUALLY.**
 
-### Step 3: Set up the agent inside Docker sandbox
+### 3️⃣ Step 3: Set up the agent inside Docker sandbox
 
 Authenticate inside the Docker sandbox before running Ralph. Run:
 
@@ -76,48 +107,16 @@ docker sandbox run claude .
 
 And follow the instructions to log in into Claude Code.
 
-> Answer "Yes" to "Bypass Permissions mode", that's the exact reason why you are using the Docker sandbox.
-
-### Step 4: Run Ralph
+👉 Answer "Yes" to `Bypass Permissions mode`, that's the exact reason why you are using the Docker sandbox.
 
-```bash
-./ralph.sh -n 50 # Run Ralph Loop with 50 iterations
-```
+> If you want to use a different agentic CLI, see [Running with a different agentic CLI](#running-with-a-different-agentic-cli).
 
-### (optional) Adjusting to your language/framework
-
-This script assumes the following are installed:
-- [Playwright](https://playwright.dev/) for e2e testing
-- [Vitest](https://vitest.dev/) for unit testing
-- [TypeScript](https://www.typescriptlang.org/) for type checking
-- [ESLint](https://eslint.org/) for linting
-- [Prettier](https://prettier.io/) for formatting
-
-I recommend using a CLI to bootstrap your project with the necessary tools and dependencies, e.g.:
+### 4️⃣ Step 4: Run Ralph
 
 ```bash
-npx create-vite@latest app --template react-ts
-# or
-npx create-next-app@latest app
-```
-
-If you must start from a blank slate, which is not recommended, you can use the following commands to install the necessary tools and dependencies:
-
-Install with:
-
-```bash
-npm i @playwright/test vitest jsdom typescript eslint prettier -D
-
-# If using React, also recommend installing:
-npm i @vitejs/plugin-react @testing-library/dom @testing-library/jest-dom @testing-library/react @testing-library/user-event -D
+./ralph.sh -n 50 # Run Ralph Loop with 50 iterations
 ```
 
---------------------------------
-
-⚠️ If you are using a different language or testing framework, please adjust `.agent/PROMPT.md` to reflect your setup, server ports and startup commands etc.
-
-⚠️ The default "mode" is "implementation". Depending on your use case, you might want to change `.agent/PROMPT.md` to a different mode, e.g. "refactor", "review", "test" etc.
-
 ## Run the loop
 
 ```bash
@@ -138,6 +137,25 @@ npm i @vitejs/plugin-react @testing-library/dom @testing-library/jest-dom @testi
 
 > NB: you might need to run `chmod +x ralph.sh` to make the script executable.
 
+> ⚠️ The default "mode" is "implementation". Depending on your use case, you might want to change `.agent/PROMPT.md` to a different mode, e.g. "refactor", "review", "test" etc.
+
+⚠️ If you want to use a different language or testing framework, see below.
+
+### (Optional) Adjusting to your language/framework
+
+This script assumes the following are installed:
+- [Playwright](https://playwright.dev/) for e2e testing
+- [Vitest](https://vitest.dev/) for unit testing
+- [TypeScript](https://www.typescriptlang.org/) for type checking
+- [ESLint](https://eslint.org/) for linting
+- [Prettier](https://prettier.io/) for formatting
+
+If you'd like to use a different language, testing framework etc. please adjust `.agent/PROMPT.md` to reflect your setup, server ports and startup commands etc.
+
+👉 The loop is controlled by this prompt, which will be sent to the agent each iteration.
+
+---------------------------------
+
 ## How It Works
 
 Each iteration, Ralph will:
@@ -189,7 +207,7 @@ The agent will check this file each iteration and if it finds any critical work,
 The `ralph.sh` script is designed to be hackable.
 It is configured to use Claude Code in a Docker sandbox by default, but with a one-liner change you can change it to use any other agentic AI CLI.
 
-Check the `ralph.sh` script around `# 👉 This is the main command loop.` for the main command loop.
+Check the `ralph.sh` script around `# This is the main command loop.` for the main command loop.
 
 > NB: skills are supported by all major agentic AI CLIs via symlinks.
 
@@ -265,8 +283,8 @@ Skills are symlinked from `.agent/skills/` to multiple locations for cross-tool
 .cursor/skills/
 ```
 
-
 ## Reference
+
 ### Playwright configuration
 
 If you are using Playwright, here is a recommended configuration:
@@ -279,28 +297,19 @@ import { defineConfig, devices } from '@playwright/test';
  */
 export default defineConfig({
   testDir: './tests',
-  /* Total timeout for the entire test run (30 minutes) */
-  globalTimeout: 30 * 60 * 1000,
-  /* Run tests in files in parallel */
   fullyParallel: true,
-  /* Fail the build on CI if you accidentally left test.only in the source code. */
+  globalTimeout: 30 * 60 * 1000,
   forbidOnly: !!process.env.CI,
-  /* Retry on failure - 2 on CI, 1 locally */
   retries: process.env.CI ? 2 : 1,
-  /* Number of parallel workers */
-  workers: process.env.CI ? 1 : 6,
-  /* Reporter to use. See https://playwright.dev/docs/test-reporters */
+  workers: process.env.CI ? 3 : 6,
   reporter: 'html',
-  /* Shared settings for all the projects below. See https://playwright.dev/docs/api/class-testoptions. */
   use: {
-    /* Base URL to use in actions like `await page.goto('')`. */
     baseURL: 'http://localhost:3000',
-
-    /* Collect trace when retrying the failed test. See https://playwright.dev/docs/trace-viewer */
     trace: 'on-first-retry',
   },
 
-  // NB: Only test in Desktop Chrome and nothing else.
+
+  // NB: only chromium will run in Docker (arm64).
   projects: [
     {
       name: 'chromium',
@@ -366,6 +375,39 @@ vi.mock('next/link', () => ({
 }))
 ```
 
+### Running with a different agentic CLI
+
+If you want to use a different agentic CLI, you can adjust the `ralph.sh` script to reflect your CLI of choice.
+
+Check the `ralph.sh` script around `# This is the main command loop.` for the main command loop.
+
+Replace `docker sandbox run claude . --` with the your favorite CLI. Remember to also update the options after the `--`.
+
+```
+docker sandbox run codex . # for Codex CLI
+docker sandbox run gemini . # for Gemini CLI
+```
+
+Docker currently supports: `claude`, `codex`, `gemini`, `cagent`, `kiro`.
+See more in [Docker's docs](https://docs.docker.com/ai/sandboxes/migration/).
+
+### Starting from scratch
+
+For AI to actually verify its implementation and for the loop to work, you need a way to verify it.
+
+To that end, at the minimum you need an end-to-end test framework and a unit test framework.
+
+For example, you can use the following commands to install Playwright and Vitest:
+
+```bash
+npm i @playwright/test vitest jsdom typescript eslint prettier -D
+
+# If using React, also recommend installing:
+npm i @vitejs/plugin-react @testing-library/dom @testing-library/jest-dom @testing-library/react @testing-library/user-event -D
+```
+
+It is recommended that you add skills for your specific language and framework. See [skills.sh](https://skills.sh) to discover existing skills.
+
 ## License
 
 MIT

package/bin/cli.js

@@ -14,16 +14,29 @@ const { copyFile, copyDir, mergeDir, exists, ensureDir } = require('./lib/copy')
 
 const PACKAGE_ROOT = path.resolve(__dirname, '..');
 const TARGET_DIR = process.cwd();
+const DEFAULT_APP_DIR = 'src';
 
 // Directories to ensure exist (created even if source doesn't exist)
 const DIRS_TO_ENSURE = [
   '.agent/history',
+  '.agent/logs',
+  '.agent/prd',
+  '.agent/screenshots',
+  '.agent/tasks',
   '.agents/skills',
   '.claude/skills',
   '.codex/skills',
   '.cursor/skills',
 ];
 
+// Directories that should have a .gitkeep file
+const DIRS_WITH_GITKEEP = [
+  '.agent/logs',
+  '.agent/prd',
+  '.agent/screenshots',
+  '.agent/tasks',
+];
+
 // Symlinks to create in skills directories (relative to each skills dir)
 const SKILL_SYMLINKS = [
   { name: 'component-refactoring', target: '../../.agent/skills/component-refactoring/' },
@@ -91,12 +104,9 @@ async function main() {
 
   // Prompt 1 — App source directory
   const appDir = await clack.text({
-    message: 'Where does your app source code live?',
-    placeholder: 'app',
-    defaultValue: 'app',
-    validate(value) {
-      if (!value.trim()) return 'Directory name cannot be empty';
-    },
+    message: 'Where does your app source code live? (e.g. src, public, etc.)',
+    placeholder: DEFAULT_APP_DIR,
+    defaultValue: DEFAULT_APP_DIR,
   });
 
   if (clack.isCancel(appDir)) {
@@ -203,38 +213,43 @@ async function main() {
     }
   }
 
-  // Ensure required directories exist and create symlinks
-  console.log();
-  display.printStep('🔧', 'Ensuring directories & symlinks');
+  // Ensure required directories exist and create markers
   for (const dir of DIRS_TO_ENSURE) {
     const dest = path.join(TARGET_DIR, dir);
     ensureDir(dest);
 
-    // Create symlinks for skills directories
-    if (dir.endsWith('/skills')) {
-      for (const link of SKILL_SYMLINKS) {
-        const linkPath = path.join(dest, link.name);
-        if (!exists(linkPath)) {
-          try {
-            fs.symlinkSync(link.target, linkPath);
-          } catch {
-            // Symlink might fail on some systems, that's ok
-          }
-        }
-      }
-      display.printSuccess(`${dir}/ (with symlinks)`);
-    } else if (dir === '.agent/history') {
-      // Create .gitignore for history folder
+    if (dir === '.agent/history') {
       const gitignorePath = path.join(dest, '.gitignore');
       if (!exists(gitignorePath)) {
         fs.writeFileSync(gitignorePath, '*\n!.gitignore\n');
       }
-      display.printSuccess(`${dir}/ (with .gitignore)`);
-    } else {
-      display.printSuccess(`${dir}/`);
+    } else if (DIRS_WITH_GITKEEP.includes(dir)) {
+      const gitkeepPath = path.join(dest, '.gitkeep');
+      if (!exists(gitkeepPath)) {
+        fs.writeFileSync(gitkeepPath, '');
+      }
     }
   }
 
+  // Create symlinks for AI config skills directories
+  console.log();
+  display.printStep('🔗', 'AI config symlinks');
+  for (const dir of DIRS_TO_ENSURE) {
+    if (!dir.endsWith('/skills')) continue;
+    const dest = path.join(TARGET_DIR, dir);
+    for (const link of SKILL_SYMLINKS) {
+      const linkPath = path.join(dest, link.name);
+      if (!exists(linkPath)) {
+        try {
+          fs.symlinkSync(link.target, linkPath);
+        } catch {
+          // Symlink might fail on some systems, that's ok
+        }
+      }
+    }
+    display.printSuccess(`${dir}/`);
+  }
+
   // Copy extra files from directories (with exclusions)
   console.log();
   display.printStep('📋', 'Extra config files');
@@ -275,6 +290,19 @@ async function main() {
     }
   }
 
+  // Replace app directory in PROMPT.md if different from default
+  if (appDir !== DEFAULT_APP_DIR) {
+    const promptFile = path.join(TARGET_DIR, '.agent/PROMPT.md');
+    if (exists(promptFile)) {
+      console.log();
+      display.printStep('📂', 'App directory');
+      const content = fs.readFileSync(promptFile, 'utf8');
+      const updated = content.replaceAll('`src`', `\`${appDir}\``);
+      fs.writeFileSync(promptFile, updated, 'utf8');
+      display.printSuccess(`.agent/PROMPT.md → ${appDir}`);
+    }
+  }
+
   // Playwright setup
   if (installPlaywright) {
     console.log();

package/bin/lib/display.js

@@ -28,11 +28,16 @@ function formatCatchphrase(text) {
   const suffix = '■▒▒▓■';
   const contentWidth = LINE_WIDTH - prefix.length - suffix.length; // 52 chars for content
 
-  const padding = contentWidth - text.length;
+  if (text.length === 0) {
+    return prefix + '▒'.repeat(contentWidth) + suffix;
+  }
+
+  const textWithSpaces = ' ' + text + ' ';
+  const padding = contentWidth - textWithSpaces.length;
   const leftPad = Math.floor(padding / 2);
   const rightPad = padding - leftPad;
 
-  return prefix + '▒'.repeat(leftPad) + text + '▒'.repeat(rightPad) + suffix;
+  return prefix + '▒'.repeat(leftPad) + textWithSpaces + '▒'.repeat(rightPad) + suffix;
 }
 
 // Catchphrases (text only, will be formatted)
@@ -77,7 +82,7 @@ ${Y}██████╗  █████╗ ██╗     ██████
 ╚══════╝ ╚═════╝  ╚═════╝ ╚═╝${R}
 
 ${D}${emptyLine}${R}
-${Y}${catchphrase}${R}
+${D}${catchphrase}${R}
 ${D}${emptyLine}${R}
 ${D}═══════════════════════════════════════════════════════════${R}
  ${C}Ralph Wiggum Loop${R} ${D}・${R} Long-running AI agents

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@pageai/ralph-loop",
-  "version": "1.1.0",
+  "version": "1.4.0",
   "publishConfig": {
     "access": "public"
   },

package/scripts/lib/display.sh

@@ -53,19 +53,19 @@ show_ralph() {
 
   local catchphrases=(
     "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm helping!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm doing my best!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm in danger!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm learnding!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒My cat's breath smells like cat food.▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒Me fail English? That's unpossible!▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒I'm asking Claude to cook pasta!▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒I found a moon rock in my nose!▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒It tastes like burning!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒When I grow up, I want to be a computer!▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm a develotron!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm helpding AI!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
-    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒I'm essential!▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm helping! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm doing my best! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm in danger! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm learnding! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒ My cat's breath smells like cat food. ▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒ Me fail English? That's unpossible! ▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒ I'm asking Claude to cook pasta! ▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒ I found a moon rock in my nose! ▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒ It tastes like burning! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒ When I grow up, I want to be a computer! ▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm a develotron! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm helpding AI! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
+    "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒ I'm essential! ▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
     "■▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒▒■▒▒▓■"
   )