Open-source skill repos
Find the public repos shaping how AI agents work.
Browse curated GitHub repos builders can install, study, adapt, compare, or integrate into real agent workflows. Each card points to the source and the trust checks to run before use.
Featured starting points
Start with repos that map cleanly to builder workflows
These are not generic AI links. Each one gives builders a concrete workflow, skill format, context pattern, or integration path to study.
Agent skill stack
GStack
Role-based CEO, design, engineering, release, docs, and QA workflows for agentic coding.
Engineering skills
Agent Skills
Production engineering skills for AI coding agents from Addy Osmani.
MCP server
GitHub MCP Server
GitHub MCP server for giving agents structured access to GitHub workflows.
Category guides
Browse by agent builder problem
Use these static guides when you want a focused comparison set instead of the full inventory.
Claude Code directory
Claude Code skills and workflows
Skill packs, directories, and workflow examples for builders shaping Claude Code into a more reliable teammate.
Open guideMCP server
MCP servers for AI agents
Model Context Protocol servers and directories that connect agents to tools, repos, memory, browsers, and business systems.
Open guidePrompt library
Prompt libraries for agent builders
Prompt references and teaching repos that help builders turn loose instructions into stricter agent workflows.
Open guideCoding agent
AI coding agents and harnesses
Open-source coding agents, CLIs, editor assistants, and harnesses for hands-on software work.
Open guideMemory and context
Memory and context tools for AI agents
Memory layers, semantic stores, and repo context tools that help agents remember useful facts without overreaching.
Open guideDesign context
Design context for AI-generated UI
Design context, rules, and examples that give coding agents stronger taste and UI constraints.
Open guideMarketing skills
Marketing and operator skills for AI agents
Domain skills for copy, CRO, analytics, SEO, positioning, and operator workflows supported by human review.
Open guideCurated comparisons
Choose between common agent repo options
These short guides compare practical first moves without pretending to rank every project in the ecosystem.
GStack vs Agent Skills
Choose between a broad operator skill stack and a focused engineering skill pack for coding agents.
Open comparisonCodex vs Claude Code vs OpenCode
Compare coding agent CLIs and workflows by review control, provider needs, and sandbox habits.
Open comparisonBest MCP servers for AI agents to evaluate first
A restrained shortlist of MCP directories and servers worth comparing before connecting agents to tools.
Open comparisonPrompt libraries vs agent skills
Decide when a public prompt is enough and when to turn the idea into a versioned agent skill.
Open comparisonCurated repo inventory
Filter by the job you need the repo to do
Use category and intent filters to find install candidates, study references, adaptation patterns, comparison sets, and integrations.
MCP directory
Awesome MCP Servers
Shared by punkpeye · punkpeye/awesome-mcp-servers
A curated directory of MCP servers that add tools and data sources to AI agents.
MCP server
GitHub MCP Server
Shared by GitHub · github/github-mcp-server
GitHub MCP server for giving agents structured access to GitHub workflows.
Agent framework
LangChain
Shared by langchain-ai · langchain-ai/langchain
Framework for building applications with LLMs, tools, retrieval, and agents.
Agent framework
LangGraph
Shared by langchain-ai · langchain-ai/langgraph
Graph-based framework for controllable, stateful agent workflows.
RAG and documents
LlamaIndex
Shared by run-llama · run-llama/llama_index
Data framework for connecting private documents and tools to LLM applications.
Workflow builder
n8n
Shared by n8n-io · n8n-io/n8n
Workflow automation platform that can connect agent steps to business systems.
Memory and context
Mem0
Shared by mem0ai · mem0ai/mem0
Memory layer for AI agents that need persistent user and task context.
Memory and context
Zep
Shared by getzep · getzep/zep
Long-term memory service for assistants with user, session, and facts context.
Memory and context
MCP Memory Service
Shared by doobidoo · doobidoo/mcp-memory-service
MCP server that gives compatible agents semantic memory capabilities.
Observability and evals
Langfuse
Shared by langfuse · langfuse/langfuse
Open-source tracing, prompt management, and evaluation for LLM applications.
Observability and evals
Phoenix
Shared by Arize-ai · Arize-ai/phoenix
Open-source observability and eval tooling for LLM, RAG, and agent systems.
Browser automation
Playwright MCP
Shared by microsoft · microsoft/playwright-mcp
MCP server that gives agents structured browser automation through Playwright.
Code graph and context
Serena
Shared by oraios · oraios/serena
Coding agent toolkit that uses language-server based semantic code understanding.
Trust check
Treat every public agent workflow as source to review, not magic to trust.
Before installing or integrating, check the license, maintainer activity, permission scope, hidden network calls, credential handling, and whether the workflow was tested on a task like yours.
Build in public with us
Have an AI agent skill, MCP server, or workflow repo worth studying?
Submit it for review. Strong fits can be added to the directory, highlighted on BuildLeanSaaS X, and included in future blog roundups for builders comparing real agent workflows.
Submit yours for X highlightBuilder learning path
Want help turning these repo ideas into working agent systems?
BuildLeanSaaS teaches builders how to evaluate public examples, design safer workflows, and ship agent-backed product systems with review loops.
Explore the skills marketplace