Category
Rulesrules, skills & MCP servers
Rules are the context files that tell an AI editor how your codebase actually works — coding conventions, architectural boundaries, the libraries you allow, the patterns you ban. A good rule file is the difference between Cursor or Claude Code guessing at your style and following it. The catch is that most rules floating around GitHub are tuned for one repo and silently assume a stack you don't run. The rules below are quality-scored: we grade them on freshness, how completely they specify the editor target (.cursorrules, AGENTS.md, CLAUDE.md), and whether they carry a real license and author. Each one is attributed to the GitHub author who wrote it, so you can read their other work before you trust their conventions. Copy a rule, paste it into your editor's config location, and your assistant inherits a battle-tested style guide in one step — no prompt engineering, no trial and error. Browse the top rules below or open the full catalog to filter by framework and AI tool.
Top rules right now
Ranked by quality score — freshness, schema completeness, and review signals. Refreshed daily.
Showing the top 24 of 31 rules.
Browse other categories
MCP Servers
Model Context Protocol servers that extend AI capabilities.
Skills
Packaged capabilities your assistant can invoke on demand.
Agents
Autonomous agents and subagents for Claude Code, Cursor, Codex, and frameworks like LangGraph or CrewAI.
Agent Teams
Multi-agent crews, roles, and orchestration graphs.
Workflows
End-to-end automations for n8n, Make, and similar tools.
Prompts
Hand-tuned system prompts and prompt libraries.
CLIs
Command-line utilities and AI-native shells.
Datasets
Training, evaluation, and retrieval datasets.
Bundles
Curated multi-item packs priced as a single drop.