NLP

Manage long-running agent sessions. Use for tracking progress in extended tasks, maintaining context across long sessions, and managing multi-step workflows.

Kyle's professional background, work context, and general preferences for R development. Use this skill to understand his expertise level and tailor explanations appropriately.

Server-Sent Events (SSE) streaming for Claude API with support for text, tool use, and extended thinking. Activate for real-time responses, stream handling, and progressive output.

Detects project tech stack, languages, frameworks, package managers, cloud platforms, and generates appropriate permissions.TRIGGERS - Keywords: tech stack, stack detection, detect stack, analyze project, project analysis, what technologies, what framework, what language, package manager, dependencies, bootstrap, setup, initialize, /claudenv, infrastructure setup, permissions, project context, environment detection.TRIGGERS - Phrases: "what stack is this", "what's this project using", "detect the technologies", "analyze this project", "set up permissions", "bootstrap infrastructure", "identify frameworks".

Context isolation for query-type MCP tools (LSP, search, database) via external CLI. Use when MCP query results consume too many context tokens.

Expert guidance on writing in Chamath Palihapitiya's communication style across 5 formats (Annual Letters, Customer Briefs, Emails, Policy Ideas Briefs, Learn with Me Presentations). Use when helping with any writing task, document creation, editing, reviewing communications, or answering questions about how to write something. Includes 10 universal principles, format-specific playbooks, internal team email guidance, templates, decision frameworks, and quality control checklists. MODULAR ARCHITECTURE: Optimized for LLM consumption with selective context loading (core + format).

Human writing standards to eliminate AI-generated text patterns. This skill should be used for any copywriting, marketing content, documentation, or professional writing to ensure output reads as genuinely human. Covers banned vocabulary, punctuation discipline, sentence variety, and tone guidelines.

Category theory for compositional computing with string diagrams, quantum circuits, and QNLP. Covers monoidal categories, functors, tensor evaluation, and practical applications in quantum natural language processing and diagrammatic reasoning.

Enterprise Clerk Authentication Platform with AI-powered modern identity architecture, Context7 integration, and intelligent user management orchestration for scalable applications

Python async/await patterns with asyncio, concurrent.futures, threading, and multiprocessing. Covers async context managers, timeouts, cancellation, common pitfalls (blocking in async, missing await, event loop issues), and choosing between async/threading/multiprocessing. Use when writing async code, debugging async issues, choosing concurrency approaches, or testing async functions.

Parse WCAG accessibility scan reports and combine with personal accessibility testing experiences to generate compelling, evidence-based violation narratives that cite both user impact and technical violations. Generates plain-text complaint narratives suitable for demand letters, legal notices, and accessibility audit reports.

Current state of Claude's memory, context, and Skills architecture as of late 2025. LOAD THIS when discussing context engineering, Skills, userMemories, memory management, context portability across surfaces, or troubleshooting why context isn't loading. Contains experimental findings and observed behaviors NOT in training data.

Parse large documents that exceed context limits into structured sections with abstracts, metadata, and hierarchies. This skill should be used when encountering documents over the context limit (typically 25k+ tokens) or when a user explicitly requests document parsing. Apply layout-aware hierarchical chunking principles to preserve semantic structure.

Discovers and finds connection details for Model Context Protocol (MCP) servers through research and documentation

Provide L1-L5 pedagogy layer reference for the AI-Native Robotics Textbook. Use when assigning layers to content, understanding layer requirements, or validating layer progression.

This skill should be used when the user asks about "MCP", "Model Context Protocol", "MCP client", "SSE client", "STDIO client", "external tools", or needs to work with MCP integrations in XSky.

This skill should be used when solving hard questions, complex architectural problems, or debugging issues that benefit from GPT-5 Pro or GPT-5.1 thinking models with large file context. Use when standard Claude analysis needs deeper reasoning or extended context windows.

Triggers when reading or writing documents exceeding 20,000 tokens (~80,000 characters or ~15,000 words). Use when encountering large files, lengthy documentation, big markdown documents, massive PRDs, architecture docs, or any content that risks context window overflow. Activates on phrases like "this document is long", "large file", "too many tokens", or when Claude detects document size approaching limits.

Autonomous browser agent for complex multi-step web tasks. WHEN: User needs complex web automation, multi-page workflows, form filling sequences, tasks requiring multiple browser actions. WHEN NOT: Simple single-action tasks like one click (use browser_click), one navigation (use browser_navigate), one extraction (use browser_extract_text).

Enhance user prompts by analyzing project context (code structure, dependencies, conventions, existing patterns). Use when users provide brief development requests that would benefit from project-specific context to generate more accurate, contextually-aware prompts.