القائمة
use when creating or updating end-to-end tests for the Seer application. This skill provides guidance on how to structure and implement e2e tests effectively.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
React Flow best practices for workflow canvas in seer-frontend. Use when modifying WorkflowCanvas, adding new block types, debugging node rendering issues, or optimizing React Flow performance.
Apply Seer design system patterns when creating or modifying UI components. Auto-triggers for badge, button, input, and form components in src/components/.
Use this when you are asked to debug production errors
Guide for adding new external integrations (OAuth or API key-based) to seer. Use when implementing new service integrations like Google, GitHub, Slack, etc.
Context about runtime credential resolution in seer. Use when understanding how OAuth tokens, resources, and secrets are resolved for tool execution.
Context about how integration metadata configures frontend display in seer. Use when understanding INTEGRATION_CONFIGS, icons, scopes, or how the dynamic UI system works.
| name | e2eTests |
| description | use when creating or updating end-to-end tests for the Seer application. This skill provides guidance on how to structure and implement e2e tests effectively. |
We use actual testing backend for our E2E playwright test with all auth config connected. and hence it is hard to guess the rendered tool/trigger/node configs and other data , since it comes via apis.
To understand the app better you can use playwright mcp to open our localy runnig frontend at http://172.17.0.1:3003 and then you can inspect the actual app and design your playwright tests by combining the observations from actual app and the codebase.Credintials for login are at .env.test file in the repo.
Use semantic locators only — always use getByRole, getByLabel, getByText, or getByTestId; never use CSS selectors, XPath, or positional selectors like nth-child.
Never use waitForTimeout — instead wait for specific UI state using await expect(locator).toBeVisible() or await page.waitForResponse('**/api/...') before proceeding.
Always use web-first assertions — use await expect(locator).toBeVisible/toHaveText/toBeEnabled() so Playwright auto-retries; never read DOM values and assert separately.
Assert UI is ready before acting — before every meaningful interaction (click, fill, submit), assert the target element is visible/enabled first to prevent race conditions with your real backend's response times.
