list_a11y_review_queue

Read the a11y REVIEW QUEUE - the per-(key, locale, surface) review items with their state. Each item: {key_uuid, key_name, namespace_slug, surface, surface_kind, locale, state (missing | draft | translated | reviewed | approved | rejected), value, is_bot, gap, base_value, screenshots[], drift, ignored, ignore_reason, reject_reason}; the envelope carries by_state / by_surface rollups. Pass state='rejected' to drive the RECTIFICATION loop: a reviewer rejected a suggestion and left a reject_reason - the retained value is the rejected text and reject_reason is the human's instruction. The agent regenerates a corrected value addressing the reason and re-proposes via set_a11y_variant (rejected is unprotected, so it returns to draft). Defaults to the production version. READ-ONLY.

Parameters

Name	Type		Description
project_uuid	string	optional	Project UUID. Required when multiple projects are configured (SONENTA_PROJECTS); defaults to the single configured project otherwise.
version_id	string	optional	Optional version UUID; defaults to the production version.
state	string	optional	Filter to one review state. Use 'rejected' for the rectification loop (items carry reject_reason). (missing, draft, translated, reviewed, approved, rejected)
locale	string	optional	Filter to one locale code (e.g. 'fr'). Use 'source' for source-locale rows.
surface	string	optional	Filter to one a11y surface. (aria_label, alt_text, screen_reader, plain_language)
namespace_uuid	string	optional	Restrict to one namespace by its UUID.
key_uuid	string	optional	Restrict to one key by its UUID.
include_ignored	boolean	optional	Include items suppressed by the remediation plan (ignored cells). Default false.
group_by_key	boolean	optional	Aggregate items by key (one entry per key with its surfaces/locales nested) instead of a flat list.
limit	integer	optional
offset	integer	optional

Example

{
  "tool": "list_a11y_review_queue",
  "arguments": {}
}