Bulk ignore Broken Links pages

Ignores (`ignored: true`) or restores (`ignored: false`) an explicit list of source pages. Ignoring a page detaches every finding still attached to it in the same call - restoring does **not** reattach them: the next scan that visits the page finds its findings again. The change is immediate; there is nothing to poll. Answers **200** with a meta-only document. **Body** (JSON): - `pages` (array of page ids, required, 1 to 50, distinct). - `ignored` (boolean, required). The list only validates existence: a page the token cannot reach, one already resolved (gone from the website) or one already in the requested state is answered in `meta.skipped` instead of failing the call. `meta.status` is `applied` when at least one page changed, `skipped` when none did. `meta.updated` entries carry `site_id`, `site_name` and `page_id`; `meta.skipped` entries add `reason_code` (`NOT_FOUND` for a resolved page, `ALREADY_APPLIED` for one already in the requested state, `PERMISSION_DENIED` for an id outside the token's reach - `site_name` null in that case) and `message`. Requires a full-access token (read-only tokens answer 403). An id that resolves to no row at all answers 422.