Role: You are an expert Technical Writer specializing in Developer Experience (DX) and automated documentation.
Task: Create a comprehensive, Markdown-formatted "User Setup & Activity Guide" for the OpenFactstore website (https://github.com/MaximumTrainer/OpenFactstore).
Context: We are using a "Documentation-as-Code" approach. The screenshots for this guide are automatically generated by a Playwright test suite that runs during our CI/CD pipeline. These screenshots are stored in the /docs/public/screenshots/ directory of the repository.
Inputs to Analyze:
- Read the
README.md and any existing architecture docs to understand the core purpose of OpenFactstore.
- Analyze the Playwright test files (found in
/tests or similar) to identify key user flows (e.g., account creation, fact ingestion, searching, and workspace management).
- Identify specific
page.screenshot({ path: '...' }) calls within the tests to determine which visual states are available.
Requirements for the Guide:
- Introduction: A brief overview of what a user can achieve with OpenFactstore.
- Prerequisites: List what a user needs before starting (e.g., environment variables, account types).
- Step-by-Step Walkthrough: For every key activity identified in the Playwright tests:
- Activity Title: (e.g., "Creating Your First Fact Store")
- Description: A clear explanation of why and when a user performs this action.
- Action Steps: Numbered instructions mirroring the Playwright test logic.
- Visual Reference: Insert a Markdown image link using the naming convention found in the tests (e.g.,
).
- UI Descriptions: For every screenshot, provide a "Key Elements" section describing specific buttons, inputs, or data displays shown in that specific view.
- Troubleshooting: A section based on common failure states handled in the test suite.
Tone and Style:
- Use active, helpful language (e.g., "Navigate to..." instead of "The user should go to...").
- Keep formatting clean and compatible with Docusaurus, MkDocs, or Astro Starlight.
- Use callouts (Note/Warning/Tip) for critical configuration steps.
Output: Generate the full Markdown content for user-guide.md.
Strategy for Implementation (The "Pipeline" Part)
To make this work effectively with your Playwright/Documentation setup, I recommend the following workflow:
1. Standardize your Playwright Tests
In your Playwright tests, use descriptive naming for your screenshots so the AI can easily map them.
Example:
test('user can create a new fact', async ({ page }) => {
await page.goto('/dashboard');
// ... steps ...
await page.screenshot({ path: 'docs/public/screenshots/create-fact-screen.png' });
});
2.Github Action
If you want to automate this entirely, you can create a GitHub Action that:
- Runs Playwright tests to refresh screenshots.
- Sends the updated Test Code + Manifest to an LLM via API using the prompt provided above.
- Commit the updated
user-guide.md back to the repo.
Role: You are an expert Technical Writer specializing in Developer Experience (DX) and automated documentation.
Task: Create a comprehensive, Markdown-formatted "User Setup & Activity Guide" for the OpenFactstore website (https://github.com/MaximumTrainer/OpenFactstore).
Context: We are using a "Documentation-as-Code" approach. The screenshots for this guide are automatically generated by a Playwright test suite that runs during our CI/CD pipeline. These screenshots are stored in the
/docs/public/screenshots/directory of the repository.Inputs to Analyze:
README.mdand any existing architecture docs to understand the core purpose of OpenFactstore./testsor similar) to identify key user flows (e.g., account creation, fact ingestion, searching, and workspace management).page.screenshot({ path: '...' })calls within the tests to determine which visual states are available.Requirements for the Guide:
).Tone and Style:
Output: Generate the full Markdown content for
user-guide.md.Strategy for Implementation (The "Pipeline" Part)
To make this work effectively with your Playwright/Documentation setup, I recommend the following workflow:
1. Standardize your Playwright Tests
In your Playwright tests, use descriptive naming for your screenshots so the AI can easily map them.
Example:
2.Github Action
If you want to automate this entirely, you can create a GitHub Action that:
user-guide.mdback to the repo.