skills-library

Optimize Three.js and WebGL performance, reduce GPU load, suggest fallback UI strategies, and maintain 60fps experience.

Agent-to-Agent (A2A) messaging system for direct communication between Cloud and Local agents. Replaces file-based handoffs with real-time messaging while maintaining vault as audit record. Supports task delegation, status updates, approval requests, and result delivery between distributed AI Employee instances. Use when coordinating multiple Claude Code agents across cloud and local environments.

Implements the core Customer Success Agent using OpenAI Agents SDK. Manages the orchestration of tool calls (search, ticket creation, etc.), conversation logic, and state management. Use when Claude needs to create or modify Customer Success Agent implementations with OpenAI Agents SDK, orchestrate multiple tools, manage conversation flows, or handle persistent state across interactions.

Generate valid AGENTS.md files for repositories according to AAIF standards. Use when Claude needs to create or update AGENTS.md files that document AI agents in a repository following standard format and best practices.

Design learning experiences for AI-native software development, integrating the Three Roles Framework (AI as Teacher/Student/Co-Worker), co-learning partnership pedagogy, and "Specs Are the New Syntax" paradigm into programming curriculum. Use this skill when educators need to prepare students for professional AI-driven development workflows, teach effective specification-first collaboration, or balance AI assistance with foundational learning goals. This skill helps create lessons that leverage AI tools appropriately while ensuring students build independent capability, bidirectional learning patterns, and ethical AI use practices. Aligned with Constitution v4.0.1.

Expert in LLM integration (Claude/GPT), RAG systems for self-learning, and WhatsApp Business API for remote approval workflows. Use when: (1) Designing prompts for Claude or GPT models with system prompts and few-shot learning, (2) Implementing multi-agent orchestration with specialized agent prompts, (3) Setting up RAG systems with vector databases (Pinecone, ChromaDB, Weaviate), (4) Implementing document chunking, embedding generation, and retrieval strategies, (5) Building WhatsApp approval workflows with interactive messages and buttons, (6) Integrating feedback loops for continuous learning, (7) Optimizing LLM API calls with error handling and retries, (8) Creating message templates for WhatsApp Business API. Provides prompt templates, RAG implementation patterns, WhatsApp message templates, and comprehensive reference documentation.

Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.

Deploy applications to Kubernetes using Argo CD GitOps workflows. Use when Claude needs to work with Argo CD for GitOps-based deployments. GitOps pattern enforced, no CI pipeline design, application manifests only.

Creates assessments with varied question types (MCQ, code-completion, debugging, projects) aligned to learning objectives with meaningful distractors based on common misconceptions. Activate when educators design quizzes, exams, or tests measuring understanding; need questions at appropriate cognitive levels (Bloom's taxonomy); want balanced cognitive distribution (60%+ non-recall); or require rubrics for open-ended questions. Generates MCQs with diagnostic distractors, code-writing prompts, debugging challenges, and project-based assessments targeting deep understanding.

Structured JSON audit logging for all agent actions with retention and traceability guarantees. Use when Claude needs to maintain comprehensive logs of all operations, track changes, ensure compliance, or provide forensic capabilities for agent activities.

Build production-ready AI microservices with Docker and streaming support. Use when the user needs to: (1) create REST endpoints that call an LLM API (OpenAI, Anthropic, or compatible), (2) stream LLM responses to clients via SSE (Server-Sent Events), (3) containerize a Node.js/Express AI service with Dockerfile and docker-compose, (4) manage secrets via environment variables with runtime validation, (5) add structured logging with per-request correlation IDs. Triggers on keywords like: AI microservice, LLM endpoint, streaming, SSE, Dockerfile, docker-compose, request ID, logging, OpenAI API, Anthropic API, containerize, AI service.

Build secure RESTful APIs with authentication, validation, and PostgreSQL integration. Use when the user needs to: (1) design CRUD endpoints with consistent response envelopes, (2) implement JWT authentication with protected routes, (3) validate and sanitize request input with Zod, (4) design PostgreSQL schemas with soft delete (deleted_at), (5) add cursor or offset-based pagination with filtering and sorting, or (6) apply rate limiting (100 req/min/IP). Triggers on keywords like: REST API, CRUD, JWT, auth, pagination, rate limit, soft delete, Express router, validation, filtering, sorting.

Implement reusable Python BaseWatcher abstractions for continuous monitoring and event emission into Obsidian vault folders. Use when building file system watchers, API monitors, or any continuous monitoring solution that outputs Markdown events to Obsidian. Triggers for tasks involving: (1) monitoring directories for file changes, (2) polling REST APIs for updates, (3) emitting structured events to Obsidian notes, (4) creating custom watchers extending BaseWatcher.

Plan, structure, and scaffold large educational books using cognitive load management, just-in-time specification, and pedagogical best practices. Use this skill when planning multi-part, multi-chapter educational works that require narrative continuity, progressive complexity, and hands-on exercises. This skill helps create cohesive learning journeys that balance foundational scaffolding with advanced independence. Activate when tasks involve structuring books, managing cognitive load across chapters, defining part-level narratives, or coordinating multi-phase content development workflows.

Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.

MCP server for browser-based payment portal automation using Playwright. Supports navigating payment sites, filling forms, drafting payments with mandatory HITL approval before execution. Never auto-approves payments. Use when Claude needs to interact with payment portals, banking websites, or any web-based financial operations requiring human oversight.

Analyze business tasks, transactions, and goals to generate executive-level weekly briefings and CEO reports with KPIs, financial summaries, and performance indicators.

MCP server enabling Claude Code to create, update, and query Google Calendar events with HITL approval for shared calendar modifications. Use when Claude needs to manage scheduling, create meetings, check availability, or automate calendar operations.

This skill should be used when content teaches patterns (skills, subagents, ADRs, PHRs, specifications) that have canonical sources elsewhere. Prevents format drift by ensuring content references and follows the authoritative format from canonical sources. Use before implementing lessons that teach platform patterns, or when reviewing content for format consistency.

Create beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.

Implements FastAPI endpoints and webhook handlers for Gmail API, Twilio WhatsApp, and the Web Support Form. Handles message normalization and initial ingestion into the system. Use when Claude needs to create or modify webhook endpoints for multi-channel message ingestion, normalize messages from different sources, or integrate with Gmail, Twilio WhatsApp, or web forms.

Generate high-quality, reasoning-driven code examples for any language or framework. Use when you need pedagogically clear, well-commented examples that teach concepts rather than just demonstrate syntax.

Validate and test code snippets in an isolated sandbox environment. Use when you need to verify correctness, catch runtime errors, or safely execute untrusted code before integrating it into a project.

Explains educational concepts in a clear, step-by-step manner with examples and analogies

Build structured conceptual frameworks that scaffold learners from foundational ideas to advanced understanding. Use when designing curricula, onboarding materials, or multi-stage learning progressions.

This skill should be used when evaluating the quality of book chapters, lessons, or educational content. It provides a systematic 6-category rubric with weighted scoring (Technical Accuracy 30%, Pedagogical Effectiveness 25%, Writing Quality 20%, Structure & Organization 15%, AI-First Teaching 10%, Constitution Compliance Pass/Fail) and multi-tier assessment (Excellent/Good/Needs Work/Insufficient). Use this during iterative drafting, after content completion, on-demand review requests, or before validation phases.

Manages the PostgreSQL database (the 'Internal CRM'). Responsible for schema migrations, customer identity resolution across channels, ticket lifecycle tracking, and vector search implementation via pgvector. Use when Claude needs to work with PostgreSQL database operations, perform schema migrations, manage customer identity resolution, track ticket lifecycles, or implement vector search functionality.

Design efficient PostgreSQL schemas and write optimized queries for analytics workloads. Use when the user needs to: (1) design relational schemas that support fast aggregation and reporting, (2) choose indexing strategies for analytics or time-series data, (3) write aggregation queries for business metrics like ROAS, CTR, or revenue, (4) filter by time windows (last 30 days, month-over-month), (5) diagnose and optimize slow PostgreSQL queries. Triggers on keywords like: schema, index, ROAS, aggregation, GROUP BY, time series, slow query, EXPLAIN, analytics, partitioning, metrics.

Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.

Deploy Docusaurus documentation site to Kubernetes. Use when Claude needs to deploy static Docusaurus sites specifically to a Kubernetes cluster. For GitHub Pages deployment, use docusaurus-deployer instead.

This skill should be used when deploying a Docusaurus site to GitHub Pages. It automates the configuration, building, and deployment process, handling GitHub Pages setup, environment configuration, and CI/CD automation. Includes local validation before GitHub Actions triggering.

Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. When Claude needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks

MCP server enabling Claude Code to draft, send, and search emails via Gmail with HITL (Human-in-the-Loop) enforcement. Use when Claude needs to interact with Gmail for email operations while maintaining human oversight and approval for sensitive actions.

Design and implement event-driven cloud architectures with Kafka topics, Dapr pub/sub, and scalable microservice patterns. Use when designing distributed systems, implementing async messaging, creating event-based workflows, or building loosely coupled services. Covers fault-tolerant, cloud-portable architectures.

Implements Kafka-based event streaming to decouple channel intake from agent processing. Manages producers, consumers, and message reliability across the processing pipeline. Use when Claude needs to implement Kafka-based event streaming, set up producers/consumers, manage message reliability, or decouple system components using event streaming.

Designs deliberate practice exercises applying evidence-based learning strategies like retrieval practice, spaced repetition, and interleaving. Activate when educators need varied exercise types (fill-in-blank, debug-this, build-from-scratch, extend-code, AI-collaborative) targeting learning objectives with appropriate difficulty progression. Creates exercise sets that apply cognitive science principles to maximize retention and skill development. Use when designing practice activities for Python concepts, creating homework assignments, generating problem sets, or evaluating exercise quality.

Scaffold and deploy FastAPI microservices with Dapr sidecars. Use when Claude needs to create stateless FastAPI microservices with Dapr integration for service invocation and pub/sub. No business logic authoring - focus on scaffolding and deployment patterns only.

Expert in developing deterministic FastAPI backends for content delivery, navigation, and rule-based quiz grading. Knowledgeable in Zero-Backend-LLM principles, ensuring no LLM calls are made in the core API routes.

Monitor local filesystem for dropped files and generate corresponding metadata/action Markdown files in Obsidian vault. Extracts full metadata including EXIF for images, text from PDFs, audio tags, video properties, and archive contents. Extends BaseWatcher framework. Use when building file inbox automation, document processing pipelines, or media organization systems.

Monitor banking transactions from CSV files and banking APIs, logging to Obsidian vault. Detects subscription patterns, flags unusual transactions, and generates financial summaries. Extends BaseWatcher framework. Use when building financial monitoring, expense tracking, subscription auditing, or automated bookkeeping systems.