Collections

This skill provides structured guidance for building AI agent systems with effective context management. It covers fundamentals, architectural patterns, and operational techniques for multi-agent coordination, memory design, and tool development. The collection helps developers avoid common pitfalls in context engineering.

This is Anthropic's official repository demonstrating Claude Skills, providing examples, templates, and the Agent Skills specification. It shows how to create folders with SKILL.md files that teach Claude specialized tasks through instructions and resources.

This Skill provides detailed guidance for working with the tldraw monorepo, a React-based infinite canvas SDK. It covers setup, essential commands, architecture, and detailed patterns for writing examples. The documentation is well-structured with clear command references and architectural explanations.

A repository management system for Claude Code skills, providing templates, verification scripts, and organizational structure for maintaining a collection of 72+ production skills. It includes sync checks, version validation, and documentation standards to ensure consistency across skills.

A meta-skill providing comprehensive guidelines for creating Claude Skills. It covers plugin structure, naming conventions, quality standards, and PR checklists. The skill teaches how to write effective skills with clear triggers, proper organization, and value-added content rather than reference dumps.

Connects Claude to Archon's knowledge base and project management system via REST API. Provides semantic search across documentation, website crawling, document upload, and hierarchical task management. Requires initial host URL configuration and API reference consultation.

A framework for creating structured prompts (PRPs) that guide AI agents through complete feature implementation workflows. It provides pre-built Claude commands, validation gates, and templates to ensure code changes follow a consistent process from planning to PR creation.

RAPTOR is an autonomous security research framework that automates offensive and defensive analysis. It provides commands for scanning, fuzzing, web testing, crash analysis, and OSS forensics. The framework includes specialized agents for root-cause analysis and evidence-backed GitHub investigations, with clear safety protocols for dangerous operations.