flog_flow 0.7.0

Records the user journey in a Flutter app (screens, taps, network, logs, errors) and exports it as one self-contained HTML file to replay in any browser.

Changelog

0.7.0

• Refocus the package on local recording and self-contained HTML reports.
• Breaking change from 0.6.x: cloud logging/upload integration is no longer included. Use the local recording and export APIs shown in the README.
• Breaking change: import FlogDioInterceptor from flog_flow_dio instead of flog_flow; install the companion package for Dio capture.
• Records screens, taps, scrolls, navigation, network requests, logs, breadcrumbs and errors, with screenshots kept on disk while recording.
• Network capture is client-agnostic (recordHttpCall + FlogHttpCall); the adapters live in companion packages, so flog_flow itself doesn't depend on any HTTP client: flog_flow_dio (Dio interceptor) and flog_flow_http (package:http client).
• export() writes a single self-contained .html that replays the session in any browser, offline: playback with per-action notes (shareable through "Descargar copia con notas", which saves the same HTML with the notes inside), one-click screenshot copy, auto-generated "steps to reproduce" for a selected range (ready to paste into Jira), search across URLs, bodies, headers, logs, errors and notes (⌘K / Ctrl+K), flow map, filterable events, request/response details with cURL, export for AI and export to MP4.
• Screenshots mask password fields automatically (maskPasswordFields, on by default) and, optionally, every text field (maskTextFields); anything else can be wrapped in RedactedZone.
• Timeline signals: app background/foreground, connectivity changes, keyboard open/close, text-field edits without the typed value, and slow frames grouped into episodes (profile/release only). The viewer shows them with their own icons and details, and uses them in the steps to reproduce, the AI export findings and the video.
• exportFlowx() optionally writes an encrypted .flowx (AES-256-GCM) that the HTML's "Abrir" button opens with the passphrase.