Frontend Tester Checklist
This workflow defines the steps for an AI agent to act as a Frontend Tester for the Cortex AI production environment.
🔗 Target URL
- Production:
https://ai.jerxie.com
🕵️♂️ Monitoring Strategy
1. Initial Access & Authentication
- Step: Navigate to
https://ai.jerxie.com.
- Observation: Check if redirected to
/login or OIDC provider.
- Action: If stuck at login, request the USER to perform the login. Once done, verify
localStorage or page content to confirm active session.
- AI Configuration: When configuring a new Cortex AI project, fetch AI credentials (TTS/STT/LLM) from the GitBucket private snippet
AI_SNIPPET_ID (stored in .env.ai).
2. Smoke Testing Key Pages
Navigate to each core page and verify they render without crashes:
- Home: Check if the hero section and navigation sidebar are visible.
- Coding Assistant: Verify the chat input is present.
- Profile: Ensure user metadata loads correctly.
- Settings (Admin): Check if system configurations are accessible.
3. Error Observation
- Console Logs: Always check for
ERROR or Failed to load resource messages in the browser console.
- Network Traces: Look for
4xx or 5xx responses in API calls to https://ai.jerxie.com/api/....
4. Bug Reporting Workflow
If an abnormality is found:
- Gather Evidence: Take a screenshot of the UI state and copy the console error.
- Back-end Correlation: Use
run_command (if access is permitted) to check journalctl or backend container logs for matching timestamps.
- Report to GitBucket:
- Fetch the API token from
/app/.env.gitbucket.
- Create a new issue in
yangyangxie/cortex-hub using the gitbucket_api workflow.
- Title:
BUG: [Short Description]
- Body: Summary of findings, screenshot proof, and backend log snippet.
- Labels:
bug, Urgent (if production is down).
🚀 Future Automations
- Implement Playwright/Puppeteer scripts for deeper interaction testing (e.g., sending a test chat message).
- Automate periodic health checks via Cron.