Troubleshooting

This page covers the most common issues people run into with Codewick and how to resolve them.

AI features not working

Symptoms: Chat returns errors, AI debugging or review won’t start, “AI unavailable” message.

Solutions:

1. Check your internet connection. AI requests require a live connection since your code is sent directly to the AI provider over HTTPS.
2. Check your token budget. Open the usage meter in the status bar. If it shows 100% or “Budget exhausted,” your monthly tokens are used up. Wait for the reset date or upgrade your plan.
3. Try a different priority mode. Switch to Cost mode in Settings > AI & Models — this uses models that may have better availability.
4. Restart the app. Quit Codewick completely and reopen it. This resets the connection to AI providers.

App won’t launch

Symptoms: Codewick icon bounces in the dock and disappears, or the app crashes on startup.

Solutions:

1. Check your macOS version. Codewick requires macOS 12 (Monterey) or later. Go to Apple menu > About This Mac to verify.
2. Reinstall the app. Download the latest version from codewick.ai, drag the old copy to Trash, then install the fresh download.
3. Check Console.app for errors. Open Console.app (in Applications > Utilities), filter by “Codewick,” and look for crash logs or error messages. Include these when reporting the issue.
4. Clear app cache. Delete ~/Library/Caches/ai.codewick.app/ and try launching again.

Browser not showing my project

Symptoms: The built-in browser panel is blank, shows an error, or doesn’t reflect your latest changes.

Solutions:

1. Check that your dev server is running. Look at the status bar — it shows the running server and port. If no server is listed, start one in the terminal (e.g., npm run dev).
2. Check for port conflicts. If another app is using the same port, your dev server may fail silently. Try a different port or kill the conflicting process.
3. Refresh manually. Click the refresh button in the browser toolbar, or press ⌘ + R while the browser panel is focused.
4. Toggle the browser panel. Close and reopen it with ⌘ + Shift + B.

Tip

If your dev server starts on a non-standard port, Codewick should auto-detect it. If it doesn’t, you can manually enter the URL in the browser panel’s address bar.

Python project not running

Symptoms: Terminal shows “python: command not found” or similar errors when running Python files.

Solutions:

1. Install Python on your system. Codewick uses your system’s Python installation. Download it from python.org if you don’t have it.
2. Check the Python path. Run which python3 in the Codewick terminal. If nothing is returned, Python isn’t in your PATH.
3. Use python3 explicitly. On macOS, the command is often python3 rather than python.

Git push/pull failing

Symptoms: Push or pull operations fail with authentication errors, “remote rejected,” or timeout messages.

Solutions:

1. Re-authenticate your Git provider. Go to Settings > Git and reconnect your GitHub or GitLab account.
2. Check repository permissions. Make sure your account has write access to the remote repository.
3. Check your internet connection. Git operations require network access to the remote.
4. Try from the terminal. Run git push directly in the Codewick terminal to see the full error output.

Slow AI responses

Symptoms: AI takes a long time to respond, or responses seem to hang.

Solutions:

1. Switch to Speed priority mode. Go to Settings > AI & Models and select Speed. This uses faster models optimized for lower latency.
2. Reduce context size. Large projects send more context to the AI. Create a .codewickignore file to exclude irrelevant directories:

   node_modules/
   dist/
   .git/
   *.min.js

3. Use @ mentions. Instead of asking broad questions, target specific files with @filename.ts to keep the context small.
4. Check your connection. A slow or unstable network will increase response times.

AI error messages explained

When an AI request fails, Codewick handles it automatically:

• Silent retry — If the first attempt fails due to a transient error, Codewick retries with the same model. You may see a brief “Retrying…” indicator.
• Escalation — If retries fail, Codewick may offer to try a “more powerful model.” This means switching to a higher-capability model that may succeed where the current one failed. Accepting this uses more tokens than the original request would have.
• Final failure — If all retries and escalation fail, you’ll see an error message with a “Try again” button. At this point, try simplifying your request or checking your connection.

Note

Escalation to a more powerful model only happens with your confirmation. It’s never automatic.

How to report a bug

If you hit an issue that isn’t covered here:

1. Click the Support button in the app (or press ⌘ + Shift + /).
2. Describe what happened, what you expected, and any error messages you saw.
3. Logs are attached automatically to help the team diagnose the issue.

You can also report bugs on the Codewick Discord in the #bugs channel.

Checking your app version

Your current Codewick version is displayed in the status bar at the bottom right of the app window. You can also find it in Settings > About. Always include your version number when reporting issues.