Captures & Storage

figranium automatically captures visual evidence of your automation runs. This is crucial for debugging, auditing, and ensuring quality.

Screenshots#

By default, every successful task execution (in agent or scrape mode) captures a final screenshot of the page.

Format: PNG
Location: public/captures/<run_id>_screenshot.png
Accessibility: View in Captures tab or via API.

Custom Screenshots#

You can trigger additional screenshots during an agent task using the screenshot action block.

Filename: <run_id>_<timestamp>_<custom_suffix>.png
Full Page: (Optional) Capture the entire scrollable area.

Video Recordings#

figranium can record full session videos (WebM format).

Enable: Set disableRecording to false in the task editor.
Location: public/captures/<run_id>.webm
Playback: View directly in the Captures tab.

Note: Recording consumes more disk space and CPU. Disable it for high-volume scraping.

figranium stores cookies and local storage state in storage_state.json.

Persistence: Cookies persist across browser restarts and task executions (unless statelessExecution is enabled).
Management: View and delete individual cookies in Settings > Data.
Clear All: Use the "Clear Storage" button in Settings to wipe all session data.

Stateless Execution#

If you enable statelessExecution in a task, it will start with a fresh browser profile (no cookies) and discard any changes upon completion. This is ideal for testing login flows or anonymous scraping.

Storage Management#

File Cleanup#

Screenshots and videos accumulate in public/captures. You can delete them:

Manually: Via the Captures tab (single delete).
Bulk: Use the API DELETE /api/data/captures/:name or clear all via Settings.

Execution Logs#

Execution metadata is stored in data/executions.json. This file grows over time. figranium automatically rotates logs (configurable via MAX_EXECUTIONS in server/constants.js, default 500) to prevent unlimited growth.