Hummingbird coordinates several optimization pipelines under one admin screen. The page cache writes static files under wp-content/wphb-cache/cache/. The Asset Optimization pipeline stores minified and combined CSS and JS bundles under wp-content/wphb-cache/assets/ and tracks the job status for each asset in wp_options under the wphb_minify_status key. The uptime monitor records check results from the WPMU DEV API in the same option store. The default Performance Test screen surfaces a Clear Cache button and a single performance score, which works on a small site and hides every interesting detail the moment one asset bundle fails to compile.

SleekView Kanban reads the same Hummingbird folders and option rows the Performance Test aggregates. Pick the Asset Optimization status field as the grouping and every asset becomes a card grouped under Queued, Optimizing, Optimized, or Failed. Card fronts can show the asset handle, the source URL, the original size, the minified size, the compression ratio, and the last update timestamp so a developer can spot bottlenecks across the asset pipeline without leaving the board view.

Dragging a card between columns calls the Hummingbird helper API. A move from Failed back to Queued resets the asset row and re-runs the minify and combine pipeline on the next cron tick. A move from Optimizing back to Queued cancels the in-flight worker. Hummingbird's automatic page cache purge on post save keeps running, so a manual board action never collides with the plugin's normal cache invalidation flow.

Hummingbird tries to do a lot under one performance score. The page cache writes static files. Asset Optimization minifies and combines CSS and JS.

The Critical CSS pipeline pulls above-the-fold styles. Uptime monitoring runs checks against the WPMU DEV API. Every one of those pipelines has a status that flows from queued to processing to optimized, with a failed branch when something goes wrong.

The Performance Test screen reduces all of that to a single score and a Clear Cache button, which works on a marketing site and goes blind the moment a third-party widget breaks the asset pipeline. A kanban board flips that around. Every asset is a card.

Every status value is a column. A glance at the board tells the team how many assets are queued, how many are stuck in the worker, how many are successfully optimized, and how many failed with a reason worth investigating. Dragging a card writes the change back through the same helpers the cron tasks call internally.

The Performance Test still exists, and still gives a one-number summary. The board exists for the rest of the time, when the team needs to see the asset pipeline the way the pipeline already sees itself.