Trello API list pagination limit change in v1.0: Managing large boards

Answer

Use the maximum allowed limit=1000 and follow the cursor returned in _links.next (the start or before parameter) until the link disappears. Insert a short pause of about 35 ms between calls to stay well under the 300‑request/10‑second rate limit, and deduplicate results by list ID to guard against occasional duplicates caused by concurrent board changes.

Confirmed Facts

• The Trello REST API v1 documents a hard maximum of limit=1000 for /boards/{id}/lists.
• Pagination is cursor‑based: each response includes a _links.next URL that encodes the needed start (or before) token for the next page.
• The rate limit is 300 requests per 10 seconds; a 35 ms interval yields ~28 requests/second, far below the threshold.

Likely Explanation

Because the endpoint can return up to 1,000 items per call, using the highest limit minimizes the number of round‑trips. The cursor mechanism guarantees that each page starts where the previous one ended, so iterating through _links.next yields the complete set without gaps, assuming the board does not change dramatically during the fetch.

Steps for This Case

1. Set limit=1000 and request GET /boards/{boardId}/lists?limit=1000 with your auth token.
2. Extract the list of objects from the response body.
3. If the response contains a _links.next field, parse its URL to obtain the next start (or before) value.
4. Wait ~35 ms, then repeat the request with the cursor value from step 3.
5. Accumulate list IDs in a Set (or deduplicate after collection) to eliminate any duplicates that could appear if a list is added/removed mid‑fetch.
6. Stop when a response lacks _links.next; the accumulated set now holds all lists on the board.

When to Ask for a Diagnostic Detail

If the board is actively modified (lists added or removed) during the retrieval, the simple deduplication approach may still miss a list that was deleted after it was read or duplicate a list that was added twice. In that case, please confirm whether the board is subject to concurrent changes; if yes, consider re‑fetching the board’s list count via /boards/{id} and comparing it to the unique IDs collected, then retrying any missing pages.

Verification (Optional)

# Example curl loop (bash)

#!/usr/bin/env bash BOARD_ID="your_board_id" TOKEN="your_token" KEY="your_key" URL="https://api.trello.com/1/boards/${BOARD_ID}/lists" LIMIT=1000 CURSOR="" ALL_IDS=() while true; do QUERY="${URL}?key=${KEY}&token=${TOKEN}&limit=${LIMIT}" if [[ -n "${CURSOR}" ]]; then QUERY+="&start=${CURSOR}" fi RESP=$(curl -sS "${QUERY}") # extract IDs (jq example) IDS=$(echo "${RESP}" | jq -r '.[].id') while read -r id; do ALL_IDS+=("$id") done <<< "$IDS" NEXT=$(echo "${RESP}" | jq -r '._links.next // empty') if [[ -z "${NEXT}" ]]; then break; fi # parse start cursor from next URL (simplistic) CURSOR=$(echo "${NEXT}" | sed -n 's/.*start=\([^&]*\).*/\1/p') sleep 0.035 # ~35 ms done printf "Unique lists: %d\n" "${#ALL_IDS[@]}"