skills.

Create distinctive, production-grade frontend interfaces with high design quality. Generates creative, polished code that avoids generic AI aesthetics.

Create new skills, modify and improve existing skills, and measure skill performance with eval-driven iteration loops.

Audit and improve CLAUDE.md files in repositories. Scans for all CLAUDE.md files, evaluates quality, outputs a graded report, then makes targeted updates.

Generate an explorable HTML report of Claude Code session usage — tokens, cache hits, subagents, skills, and expensive prompts — from local transcripts.

Analyze a codebase and recommend Claude Code automations — hooks, subagents, skills, plugins, and MCP servers — tailored to your stack.

Deep teaching skill. Reads any provided artifact (PDF, URL, paper, repo, code) and teaches every concept in two modes — a 12-year-old explanation and a technical deep-dive — for maximum retention.

Activates a terse, caveman-style response mode that strips filler, articles, hedging, and pleasantries while preserving all technical substance. Three intensity levels: lite, full, ultra.

Tailor an existing LaTeX resume to a specific job description without inventing facts or changing the document's size. Preserves macros, bullet counts, and page length.

Write a targeted, story-driven cover letter from a job description and a resume. Maps one or two real projects to the role's actual problem — no clichés, no padding.