Bulkhead visualization pattern
Skill kjuhwa/skills-hub/skills/design/bulkhead-visualization-pattern
Self-correcting knowledge corpus for Claude Code — 9 stable shape clusters, bias-correction pipeline baked into contribution flow. 47 papers, 45 techniques, 1.1k skills.
npx -y skills add kjuhwa/skills-hub --skill bulkhead-visualization-patternAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Visual layout conventions for rendering isolated resource pools side-by-side with saturation indicators
SKILL.md
1.7 KB, 305 tokens by cl100k_base, as published. Nobody here has run it
bulkhead-visualization-pattern
Bulkhead visualizations should present each partition as a distinct, bordered container arranged horizontally or in a grid, never merged into a single aggregate view. The core visual metaphor borrows from ship compartments: each pool gets its own "chamber" with a clear capacity ceiling (e.g., 10 slots rendered as 10 cells or a filled bar), and in-flight requests fill the chamber from the bottom up. Color-code by saturation thresholds — green (0-60%), amber (60-90%), red (90-100%) — so viewers instantly spot which partition is drowning without reading numbers.
Cross-partition comparison requires a shared y-axis scale across all bulkhead panels; otherwise a small pool at 9/10 looks identical to a large pool at 90/100 and the isolation benefit is invisible. Always render a "rejected/queued" overflow indicator outside the chamber boundary (e.g., stacked red dots spilling off the side) to show that back-pressure is kicking in on one partition while others remain healthy — this is the whole point of the pattern and must be the most prominent signal.
Include a timeline or sparkline strip beneath each chamber showing recent saturation history (last 30-60s). This makes it obvious when one bulkhead has been pegged while neighbors idle, which is the key insight the pattern teaches. Avoid aggregating metrics across bulkheads in headline numbers — show per-partition latency p50/p99 separately, because averaging defeats the isolation narrative.