mcp-builder
jnmetacode
Build MCP servers with tools AI can use reliably.
MCP development
AI agents & skills skills for AI agents
This is where agents get better at being agents: building them on platforms such as Microsoft Foundry, connecting them through MCP servers and giving them memory. Next come Skills about Skills, which write, find and manage them, then prompt engineering and evaluation of AI features.
Sort by
Most installed
840 Skills
Showing Skills in AI agents & skills.- 936
932autoresearch
orchestra-research
Turn a research question into tested findings.
AI research
929agent-evaluation
mlflow
Find where your LLM agent's answers fall short.
Agent evaluation
922agent-manager-skill
sickn33
Keep parallel local agents and their work in view.
Agent orchestration
- 918
crewai-multi-agent
orchestra-research
Build AI agent teams that share complex work.
AI agent development
- 903
dspy
orchestra-research
Build AI pipelines with prompts that improve.
AI development
- 902
axolotl
orchestra-research
Fine-tune language models with Axolotl guidance.
LLM fine-tuning
900agents
calesthio
Build voice assistants with natural conversations.
Voice AI development
- 900
langchain
orchestra-research
Build LangChain apps that answer from your docs.
LLM application development
+2 - 897
sglang
orchestra-research
Serve language models with structured outputs.
LLM inference
SG 
893create-task
harbor-framework
Build a Harbor benchmark task you can verify.
Agent evaluation
893generate-synthetic-data
ai-evals-course
Realistic test queries for your LLM pipeline.
LLM evaluation
- 893
slime-rl-training
orchestra-research
Train language models with reinforcement learning.
LLM post-training
- 892
llamaindex
orchestra-research
Build apps that answer questions from your docs.
LLM application development
891agent-skills-creator
mblode
Build skills that earn their place in your agent.
Skill authoring
- 890
guidance
orchestra-research
Make your LLM outputs fit the format you need.
LLM development
+1 - 889
implementing-llms-litgpt
orchestra-research
Build and fine-tune language models with LitGPT.
Language model development
- 888
nemo-guardrails
orchestra-research
Add runtime safety checks to your LLM app.
LLM safety
- 887
rwkv-architecture
orchestra-research
Build streaming text applications with RWKV.
Language modeling
- 884
constitutional-ai
orchestra-research
Train safer models with AI-generated feedback.
AI safety
- 884
pinecone
orchestra-research
Build AI search with managed vector storage.
AI search
882subagent-creator
tech-leads-club
Focused AI helpers for complex coding work.
Agent development
881omni-mcp
diegosouzapw
Connection guidance for OmniRoute's MCP tools.
MCP integration
+1- 879
build-review-interface
ai-evals-course
Review LLM traces with a custom labeling page.
LLM evaluation
A folder of instructions, and sometimes scripts, that an agent loads when a task calls for it. It teaches the agent a way of working it would not follow on its own.
Yes. Skill authoring Skills guide an agent through writing a new Skill, testing it on real tasks and improving it from what went wrong.
The Model Context Protocol, a standard way for an agent to reach outside tools and data. MCP Skills help build servers that expose a service to agents this way.