Quickstart Screenshots¶
This directory contains screenshots for the Quickstart Guide.
Automated Screenshots (via Playwright)¶
These screenshots are generated automatically by running:
export PRELOOP_USERNAME='your-username'
export PRELOOP_PASSWORD='your-password'
export OPENAI_API_KEY='your-openai-key'
export PRELOOP_BASE_URL='https://preloop.ai' # Optional
python docs/scripts/capture_quickstart_screenshots.py
Generated screenshots:
- ✅ signup.png - Sign up page
- ✅ signup2.png - Stripe checkout (not currently used)
- ✅ dashboard.png - Dashboard overview
- ✅ api-key-create.png - Create API key dialog
- ✅ api-key-success.png - API key created successfully
- ✅ add-mcp-server.png - Add MCP server dialog
- ✅ mcp-server-added.png - MCP server with tools
- ✅ tools-list-external.png - Tools list
- ✅ tool-config-approval-toggle.png - Enable approval
- ✅ approval-workflow-create.png - Create approval workflow
- ✅ add-ai-model-dialog.png - Add AI model dialog
- ✅ ai-model-created.png - AI model created
- ✅ flow-create-form.png - Create flow form
- ✅ flow-created-success.png - Flow created
- ✅ flow-test-run-dialog.png - Test run dialog
- ✅ flow-execution-started.png - Execution started
- ✅ flow-execution-waiting-approval.png - Waiting for approval
Manual Screenshots (TODO)¶
These screenshots need to be captured manually:
1. approval-email-from-flow.png¶
What: Email notification showing approval request
How to capture:
1. Run a flow that triggers an approval
2. Check your email for "Approval Required: pay"
3. Take screenshot of the email showing:
- Subject line
- Approval details (tool name, parameters)
- Approve/Decline buttons
4. Save as approval-email-from-flow.png
Recommended dimensions: 800x600px Recommended format: PNG with email client chrome visible
2. flow-execution-completed.png¶
What: Flow execution page after successful approval and completion
How to capture:
1. Trigger a test flow with approval required
2. Approve the request via email or dashboard
3. Go to flow execution page
4. Wait for execution to complete
5. Screenshot showing:
- Execution status: "Completed"
- Execution logs/timeline
- Tool calls (including approved pay call)
- Final result
6. Save as flow-execution-completed.png
Recommended dimensions: 1920x1080px (full page) Recommended format: PNG
3. webhook-url-section.png (Optional)¶
What: Flow details page showing webhook URL
How to capture:
1. Navigate to a created flow's details page
2. Scroll to "Webhook URL" section
3. Screenshot showing:
- Webhook URL
- Copy button
- Usage example
4. Save as webhook-url-section.png
Recommended dimensions: 1200x400px (section only) Recommended format: PNG
Optional Enhancement Screenshots¶
These would enhance the guide but are not critical:
- Claude Code integration - Terminal showing Claude Code connected to Preloop MCP
- Mobile approval - iPhone/Apple Watch showing approval notification
- Approval dashboard - Web dashboard approval interface
- Slack notification - Slack message with approval request (Enterprise)
Updating Screenshots¶
To regenerate all automated screenshots:
# Set up environment
export PRELOOP_USERNAME='your-username'
export PRELOOP_PASSWORD='your-password'
export OPENAI_API_KEY='your-openai-key'
export PRELOOP_SIGN_UP=false # Set to true for fresh signup flow
# Run script
python docs/scripts/capture_quickstart_screenshots.py
Note: The script requires:
- Valid Preloop account credentials
- OpenAI API key for AI model creation
- Playwright browser installed (playwright install chromium)
Screenshot Guidelines¶
When capturing screenshots: 1. Resolution: Use at least 1920x1080px for full-page screenshots 2. Format: PNG with transparency where applicable 3. Cropping: Crop to relevant content, include context 4. Annotations: No annotations - keep original UI 5. Privacy: Blur/redact any sensitive information (API keys, emails, etc.) 6. Consistency: Use same theme/browser across all screenshots 7. Quality: Retina/2x resolution preferred for clarity
Testing¶
After updating screenshots, verify: 1. All images load correctly in the quickstart guide 2. Image paths are correct (relative to guide location) 3. Image captions are descriptive 4. Images display at appropriate size in rendered docs 5. No sensitive information is visible
Build and preview docs: