LLM & Agents

A comprehensive security skill that integrates with Secure Flow to help AI coding agents write secure code, perform security reviews, and implement security best practices. Use this skill when writing, reviewing, or modifying code to ensure secure-by-default practices are followed.

LangGraph supervisor-worker pattern. Use when building central coordinator agents that route to specialized workers, implementing round-robin or priority-based agent dispatch.

Automated quality checks and enforcement for code commits. Use when validating code quality, running pre-commit checks, ensuring test coverage, or enforcing coding standards before commits and pushes.

LangGraph human-in-the-loop patterns. Use when implementing approval workflows, manual review gates, user feedback integration, or interactive agent supervision.

Avoid common mistakes and debug issues in PydanticAI agents. Use when encountering errors, unexpected behavior, or when reviewing agent implementations.

Guides architectural decisions for LangGraph applications. Use when deciding between LangGraph vs alternatives, choosing state management strategies, designing multi-agent systems, or selecting persistence and streaming approaches.

Guidelines for multi-agent AI and learning projects with lesson-based structures. Activate when working with AI learning projects, experimental directories like .spec/, lessons/ directories, STATUS.md progress tracking, or structured learning curricula with multiple modules or lessons.

Advanced React Flow patterns for complex use cases. Use when implementing sub-flows, custom connection lines, programmatic layouts, drag-and-drop, undo/redo, or complex state synchronization.

Comprehensive codebase understanding through structured documentation and guided learning. Use when Claude needs to help a user understand how a project works by exploring and documenting the codebase architecture, concepts, and modules in a .learning/ directory, then guiding them through understanding the code with adaptive explanations. Triggered by requests like "help me understand this project", "document the architecture", "explain how this codebase works", or "guide me through the code".

Pattern for automated testing with GitHub issue creation and Claude Code auto-fixing. Creates Test → Fail → Issue → Fix → Repeat cycle until tests pass.

Testing workflow patterns and quality standards. Activate when working with tests, test files, test directories, code quality tools, coverage reports, or testing tasks. Includes zero-warnings policy, targeted testing during development, mocking patterns, and best practices across languages.

Use when execute command encounters errors - diagnostic guide for phase failures, parallel agent failures, merge conflicts, worktree issues, and recovery strategies

Implements agents using Deep Agents. Use when building agents with create_deep_agent, configuring backends, defining subagents, adding middleware, or setting up human-in-the-loop workflows.

Helps choose the right Claude Code extension mechanism (slash commands, skills, subagents, or hooks) for a given use case. Use when discussing how to extend Claude Code, automate workflows, or implement custom functionality.

GPT-5 prompting best practices for creating AI agent prompts in OpenPromo codebase when using ai sdk or agents, or in general writing llm prompts.

Design code for testability and AI/LLM ergonomics with explicit contracts and observable patterns. Use when optimizing code for AI tools, improving testability, or making codebases LLM-friendly.

This skill should be used when the user asks to "create a plugin", "scaffold a plugin", "understand plugin structure", "organize plugin components", "set up plugin.json", "use ${CLAUDE_PLUGIN_ROOT}", "add commands/agents/skills/hooks", "configure auto-discovery", or needs guidance on plugin directory layout, manifest configuration, component organization, file naming conventions, or Claude Code plugin architecture best practices.

Auto-loads when CLAUDE_CODE_TEAM_NAME environment variable is set. Provides guidance for Claude Code teammates working in swarms - identity awareness, communication patterns, task management, and coordination protocols. Use when you are a spawned teammate (not the team lead).

Creates jaw-dropping, production-ready frontend interfaces AND delivers perfectly matched real photos (Unsplash/Pexels direct links) OR flawless custom image-generation prompts for hero images, backgrounds, and illustrations. Zero AI slop, zero fake URLs.

World-class prompt engineering skill for LLM optimization, prompt patterns, structured outputs, and AI product development. Expertise in Claude, GPT-4, prompt design patterns, few-shot learning, chain-of-thought, and AI evaluation. Includes RAG optimization, agent design, and LLM system architecture. Use when building AI products, optimizing LLM performance, designing agentic systems, or implementing advanced prompting techniques.