Trello board recovery: JSON fidelity with scripted restore versus CSV simplicity with manual reconstruction for sub-four-hour RTO

Goal

Establish a board-level backup strategy that can restore a corrupted or maliciously edited board within a four-hour recovery time objective (RTO) while preserving checklists, comments, custom fields, and attachment metadata.

Constraints and uncertainty

Trello's API v2 supports automated JSON exports that capture full board fidelity, but restoration requires a multi-step scripted sequence: recreate the board, lists, cards, then checklists, attachments, and custom field items in dependency order. Rate limits (300 requests per 10 seconds per token) make large-board restores slow and prone to partial failure. CSV exports are only available through the UI and omit checklists, comments, custom fields, and attachment records, but manual import via the UI is straightforward for basic card data. Custom field definitions are workspace-scoped; restoring a JSON export to a different workspace requires manual remapping of field definitions, which the API does not automate. Trello Enterprise provides 30-day deleted-board recovery but no automated point-in-time restore for individual boards.

Questions

1. For a 2,000-card board with extensive checklists and custom fields, can a scripted JSON restore reliably complete within four hours given current rate limits, or does the multi-step dependency chain inherently exceed that window?
2. If CSV is chosen for speed, what is the realistic manual effort to reconstruct checklists, comments, and custom fields for a board of that size, and are there third-party tools that reduce this gap?
3. When restoring to a different workspace, does the custom field remapping effort negate the fidelity advantage of JSON, making CSV plus manual field recreation the more predictable path?