Optimised transient storage: generation progress queue now uses two flat integer arrays (post_queue, term_queue) instead of an array of tagged {type,id} objects, reducing serialised transient size by ~75% on large sites (e.g. ~50KB instead of ~200KB for a 4,000-item site).
Fixed: the “done” check now correctly detects when both post_queue and term_queue are empty.
0.6.0
Renamed plugin (display name, slug, text domain, and full internal code prefix) to “HS3Dev AI Content Index for llms.txt” per WordPress.org Plugin Review Team naming guidance, to place the “llms” term correctly at the end of the name and avoid trademark/category confusion.
Renamed internal code prefix from llmsg_/LLMS_Generator_ to hs3daci_/HS3DACI_, since “llms” was flagged as too common a word to safely function as a collision-safe prefix.
Fixed Tested up to value (was referencing a WordPress version that does not exist).
Added plugin owner to Contributors list.
0.5.1
Renamed slug/text-domain to hs3dev-llms-generator
Fixed admin asset loading after rename
Hardened analytics database queries (prepared statements with %i identifier placeholder)
Fixed invalid “Requires at least: 7.0” value and other readme metadata issues
General WordPress coding standards cleanup
0.5.0
Added description toggle for llms.txt output (full vs minimal mode)
Improved flexibility for SEO + AI crawler output control
0.4.0
Added multilingual support (WPML/Polylang)
Added compatibility mode for page builders
Added multisite support and role control
0.3.0
Added batch processing system for large sites
Added conflict detection for physical llms.txt files
