Skill UI

This skill analyzes the predictable ways language models lose performance as context length increases. It details core patterns, including 'lost-in-middle,' context poisoning, and distraction. Use this knowledge to diagnose unpredictable agent failures, design highly reliable systems for long document processing, and implement advanced context engineering strategies.

A comprehensive guide to understanding the complete state (context) available to Large Language Models during inference. Covers core components like System Prompts, Tool Definitions, Message History, and Retrieval Augmented Generation (RAG). Learning to engineer context is crucial for building robust, efficient, and reliable AI agent systems by managing the attention budget and ensuring high-signal information delivery.

Provides comprehensive guidance on building robust evaluation frameworks for complex agent systems. It addresses the inherent challenges of non-determinism and multiple valid paths, enabling systematic performance testing, validating context engineering choices, and measuring continuous improvements across multiple quality dimensions (e.g., factual accuracy, tool efficiency).

A comprehensive guide to establishing production-ready coding standards and context for the Windsurf AI environment. Learn how to utilize root-level rules (.windsurfrules), granular workspace rules, Model Context Protocol (MCP) server integration, and advanced prompt engineering to guide the AI and maintain consistent code quality across projects.

Helps product managers diagnose whether AI workflows rely on indiscriminate context stuffing or deliberate context engineering, then apply bounded domains, retrieval, and Research→Plan→Reset→Implement practices to improve reliability and output quality.

This skill provides comprehensive techniques for optimizing prompts used with any Large Language Model (LLM). It teaches proven methodologies—including Role definition (R), Context setting (S), Constraints (C), Instructions (I), and Templates (T)—to transform vague inputs into precision-engineered instructions. Users can reliably achieve structured JSON output, implement Chain-of-Thought reasoning, reduce hallucinations, and minimize token usage, ensuring consistent and high-quality results across different AI platforms.

Guides product managers from discovery notes to an aligned PRD by orchestrating problem framing, user context, solution definition, and success criteria so engineering and stakeholders share a clear source of truth.

This guide outlines best practices for creating high-quality, structured pull requests (PRs) using Git and GitHub CLI. It ensures that every PR description captures not only what changed, but also the 'why' and necessary context for reviewers, significantly improving the code review process and adherence to engineering standards.

This utility converts a single, high-level objective into a comprehensive, step-by-step build plan suitable for complex, multi-agent, or multi-session engineering projects. It automatically handles dependency mapping, detects parallel steps, and structures each step with self-contained context briefs, ensuring any new agent can execute it without prior knowledge. Ideal for large-scale refactoring or feature development spanning multiple PRs.

A comprehensive collection of six expert-level, structured workflows designed to elevate AI assistance from ad-hoc querying to repeatable, production-grade engineering practices. Users can evaluate prompts, design RAG architectures, plan context budgets, run security audits, build eval harnesses, and coach product strategy.

This system deploys advanced AI, NLP, and behavioral analytics to detect sophisticated Business Email Compromise (BEC) attacks. Unlike traditional filters that rely on malicious links, this solution analyzes writing style, tone, grammatical patterns, and contextual anomalies to identify pure social engineering impersonation, providing a robust defense layer against advanced phishing and fraud.

This skill manages comprehensive software development life cycle operations—including issue triage, PR reviews, version releases, bug fixing, and marketing content generation. It operates using a parallel "agent army" of dynamic agents (10-20 per task), requiring the user to adopt an Engineering Manager (EM) mindset. Key principles involve rigorous citation checking (anti-hallucination) and applying multiple roles (Product Owner, Open Source Contributor, Architect) simultaneously to ensure robust, safe, and user-centric outcomes across all platforms.