From clarity
Run Clarity's Doctor — scan the past 30 days of Claude Code token usage, find problems, recommend fixes. Writes .clarity/doctor-report.md in the current directory.
How this skill is triggered — by the user, by Claude, or both
Slash command
/clarity:clarity-doctorThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Read-only diagnostic. Scans `~/.claude/projects/**/*.jsonl` for all recent sessions across all projects, attributes cost using Opus 4.7 list prices (cache_read weighted at 0.1x, not 1x — the common mistake in DIY analyzers), surfaces the top expensive sessions and top projects, and flags 2-3 concrete fix recommendations.
Read-only diagnostic. Scans ~/.claude/projects/**/*.jsonl for all recent sessions across all projects, attributes cost using Opus 4.7 list prices (cache_read weighted at 0.1x, not 1x — the common mistake in DIY analyzers), surfaces the top expensive sessions and top projects, and flags 2-3 concrete fix recommendations.
$1 — number of days to analyze (default: 30)
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/analyze.py" --since-days "${1:-30}" --out .clarity/doctor-report.md
Then Read .clarity/doctor-report.md and summarize for the user in 3-5 bullet points:
ENABLE_PROMPT_CACHING_1H=1).claude/ structure for the hot project — see case study)Builds accessible UIs with shadcn/ui components on Radix UI + Tailwind CSS, plus canvas visuals. For React apps (Next.js, Vite, Remix, Astro), design systems, responsive layouts, themes, dark mode, prototypes.
npx claudepluginhub robotlearning123/clarity --plugin clarity