Skill UI

This skill guides users through advanced prompt optimization techniques for Large Language Models (LLMs). It functions as an expert prompt engineer coach, applying methodologies like Constitutional AI and Chain-of-Thought reasoning. Use it to transform basic, raw instructions into highly effective, production-ready prompts, significantly improving accuracy, reducing hallucinations, and optimizing token usage for cost efficiency.

Guides comprehensive evaluation of LLM applications with automated metrics, human reviews, and LLM-as-judge patterns to benchmark prompts, detect regressions, and validate deployments.

Guides implementation of Mistral AI chat completions, streaming responses, multi-turn history tracking, and configurable prompts for conversational interfaces and text generation.

Optimizes Mistral AI spending by selecting models per task, estimating token costs, enforcing daily/monthly budgets, monitoring usage, and tightening prompts—ideal for teams watching billing, cutting API spend, and triggering budget alerts.

Guides teams on reducing latency and improving throughput when integrating Mistral AI by selecting low-latency models, enabling streaming, caching deterministic requests, trimming prompts, and managing request concurrency.

Guides teams through securing Mistral AI integrations by managing API keys, sanitizing prompts and outputs, logging safely, and rotating secrets to reduce exposure when deploying LLM workloads.

Recommends suitable Nano Banana Pro image generation prompts from a 6000+ library, guiding users through direct image requests or illustrating existing articles, videos, or podcasts with matched prompt templates.

Guides tuning OpenEvidence clinical queries through streamlined prompts, dynamic caching, connection pooling, batching, and monitoring to meet latency targets and safeguard clinical response times.

Optimizes prompts for LLMs to cut token usage, lower costs, and boost response quality by analyzing, condensing, and rewriting instructions; useful when teams need faster, cheaper, clearer AI conversations.

Guides orchestrating multiple Task tool subagents with run_in_background for concurrent work, covering when to trigger parallelism, how to craft prompts, launch in one message, synchronize results, and integrate TodoWrite status updates.

Phoenix is an open-source observability platform for tracing, evaluating, and monitoring LLM applications; use it to debug prompts, run dataset evaluations, build experiment pipelines, and watch production inference in real time.

Build single-file HTML playgrounds with controls, live preview, and prompt output so users can explore large visual or structural spaces and copy actionable prompts without leaving the browser.