ContextAlt turns WordPress 7 media libraries into review queues. It flags deterministic alt-text issues, shows image usage, and builds bounded page context for drafts through the site’s AI Connector. AI never publishes: each draft needs a human apply decision, stale suggestions are blocked, and changes roll back safely. Free is GPL, stores no provider keys, and has no telemetry. Pro adds resumable batches, schedules, advanced integrations, complete history, and team workflows.
I built ContextAlt after noticing that bulk alt-text generators optimize for producing output, while editors still need context, review, and a safe way back. The constraint I kept throughout the build was simple: AI may draft; it never gets the final say. ContextAlt separates audit, context gathering, generation, review, apply, and rollback, and uses source/context hashes so an old suggestion cannot overwrite newer work. The Free plugin is GPL and available from GitHub while WordPress.org reviews the directory submission. I’d love feedback from agencies and content teams with large or multilingual media libraries—especially on the issue filters and context preview.
Report
Reviewing images in their actual page context feels much more useful than checking them one by one. Has that changed the kinds of issues users typically catch?
Report
The "no publish without human apply" guardrail is a genuinely thoughtful touch, makes me trust the AI suggestions way more. Surprised how cleanly the rollback works when you swap alt text in bulk.
Report
The rollback safety on each draft is a nice touch, makes me way more comfortable letting the AI Connector handle bulk alt-text cleanup on my older posts.
Report
The rollback safety combined with blocking stale suggestions is a thoughtful touch—clearly built by people who've actually watched an AI quietly wreck a media library at 2am.
Report
No reviews yetBe the first to leave a review for ContextAlt
Reviewing images in their actual page context feels much more useful than checking them one by one. Has that changed the kinds of issues users typically catch?
The "no publish without human apply" guardrail is a genuinely thoughtful touch, makes me trust the AI suggestions way more. Surprised how cleanly the rollback works when you swap alt text in bulk.
The rollback safety on each draft is a nice touch, makes me way more comfortable letting the AI Connector handle bulk alt-text cleanup on my older posts.
The rollback safety combined with blocking stale suggestions is a thoughtful touch—clearly built by people who've actually watched an AI quietly wreck a media library at 2am.