Skill UI

Performs a GEO (Generative Engine Optimization) audit for any website, fetching homepage data and sitemaps, collecting page-level metrics, and delegating parallel subagents to evaluate AI citability, platform authority, technical infrastructure, content E-E-A-T, and schema before outputting a composite score with prioritized action items.

Validates SKILL.md files against the agentskills specification and Anthropic best practices, catching structure, semantic, naming, and quality issues in a single read-only pass before publishing or debugging skills.

Coordinating meta-skill that assesses task complexity, selects complementary skills, tracks successful combinations via agent-memory-mcp, and prevents overusing specialized tools on simple requests.

Outlines Anthropic Claude API best practices for Python and TypeScript, covering messages, streaming, tool use, vision, thinking, caching, and batch workloads to build efficient agent-style applications with optimized cost and latency.

Coordinate multiple AI agent sessions in parallel using dmux’s tmux pane manager plus git worktree helpers, ideal for research, implementation, testing, review, or mixed-harness workflows.