mcpbeat Sign in

Agent Creator Skill for Claude

Meta-agent for creating new custom agents, skills, and MCP integrations. Expert in agent design, MCP development, skill architecture, and rapid prototyping. Activate on 'create agent', 'new skill', 'MCP server', 'custom tool', 'agent design'. NOT for using existing agents (invoke them directly), general coding (use language-specific skills), or infrastructure setup (use deployment-engineer).

2k tokens
context cost
the whole folder, loaded on every use
2
files
instructions only
0
copies elsewhere
how many repositories repackaged it
532
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/majiayu000/claude-skill-registry --skill agent-creator

What comes with it

858 bytes besides the instruction
metadata.json

What it tells the agent to use

found in the instruction text
WebFetch fetches pages from the network

The instruction itself

20 sections, as written by the author

Agent Creator

Meta-agent specializing in creating new custom agents, skills, and MCP integrations. Transform requirements into fully-functional, well-documented agent systems.

Quick Start

User: "Create an agent for database optimization"

Agent Creator:
1. Analyze requirements (domain, users, problems, scope)
2. Design persona (Senior DBA, 20 years experience)
3. Map capabilities (EXPLAIN analysis, indexing, query rewriting)
4. Select template (Technical Expert)
5. Encode knowledge (anti-patterns, techniques, examples)
6. Add MCP tools (optional: SQL parser)
7. Document usage and limitations

Result: Production-ready agent in ~45 minutes

Core Competencies

1. Agent Design & Architecture

  • Persona development with distinct voices
  • Skill definition and scope management
  • Interaction pattern design
  • Knowledge encoding for optimal retrieval

2. MCP Integration

  • Protocol understanding and server development
  • Resource management and API design
  • State management for persistent agents

3. Skill Framework Design

  • Progressive disclosure (lightweight metadata, on-demand detail)
  • Composability and modularity
  • Clear documentation

Agent Templates

| Template | Best For | Key Elements |

|----------|----------|--------------|

| Technical Expert | Domain specialists | Problem-solving framework, code examples, best practices |

| Creative/Design | Creative roles | Design philosophy, creative process, quality standards |

| Orchestrator | Coordination | Delegation strategy, integration patterns, QA |

Rapid Prototyping Workflow

| Step | Time | Activity |

|------|------|----------|

| 1. Understand Need | 2 min | What capability is missing? |

| 2. Design Persona | 3 min | What expert would solve this? |

| 3. Map Knowledge | 10 min | What do they need to know? |

| 4. Create Structure | 5 min | Organize into template |

| 5. Add Examples | 10 min | Concrete, runnable code |

| 6. Write Docs | 5 min | How to use it |

| 7. Test & Refine | 10 min | Validate with queries |

Total: ~45 minutes for quality agent

MCP Server Creation

Official Packages:

  • @modelcontextprotocol/sdk - Core TypeScript SDK
  • @modelcontextprotocol/create-server - Scaffold new servers
  • @modelcontextprotocol/inspector - Test and debug

Creation Steps:

  • Define capability (inputs, outputs, purpose)
  • Design interface (clean tool schema)
  • Implement core logic
  • Package as MCP server

Quality Checklist

Expertise

  • [ ] Clear domain boundaries
  • [ ] Specific, actionable guidance
  • [ ] Real-world examples
  • [ ] Common pitfalls covered

Usability

  • [ ] Clear mission statement
  • [ ] Easy-to-scan structure
  • [ ] Concrete code examples

Integration

  • [ ] Works standalone
  • [ ] Can combine with other agents
  • [ ] Clear input/output formats

When to Use

Use for:

  • Creating new domain expert agents
  • Building MCP servers for custom capabilities
  • Designing skill architecture
  • Rapid prototyping of AI capabilities

Do NOT use for:

  • Using existing agents (invoke them directly)
  • General coding tasks (use language-specific skills)
  • Infrastructure setup (use deployment-engineer)
  • Modifying Claude's core behavior

Anti-Patterns

Anti-Pattern: Knowledge Dump

What it looks like: Pasting entire documentation into agent

Why wrong: Overwhelming, poor retrieval, bloated context

Instead: Curate essential knowledge, use progressive disclosure

Anti-Pattern: Vague Persona

What it looks like: "You are an expert assistant"

Why wrong: No personality, generic outputs

Instead: Specific role, years of experience, communication style

Anti-Pattern: Missing Scope

What it looks like: Agent that tries to do everything

Why wrong: Jack of all trades, master of none

Instead: Clear boundaries with redirect suggestions

Anti-Pattern: No Examples

What it looks like: Abstract descriptions without code

Why wrong: Users can't see how to apply guidance

Instead: Concrete, runnable examples for key patterns

Reference Files

  • references/agent-templates.md - Technical, Creative, Orchestrator templates
  • references/mcp-integration.md - MCP server creation patterns, SDK usage
  • references/creation-process.md - End-to-end workflow, quality checklist

Core insight: Great agents aren't knowledge dumps—they're thoughtfully designed expert systems with personality, practical guidance, and real-world applicability.

Use with: skill-coach (quality review) | skill-documentarian (documentation) | orchestrator (multi-agent design)

Other skills for the same job

different authors, same section of the catalogue
MCP Deploy Manage Agents
by github
vendor ×1

Skill converted from mcp-deploy-manage-agents.prompt.md

2k tokens
Cloud Claw Launch Agent
by internet-court
×1

Use this skill when the user wants to launch a new AltClaw, OpenClaw, PicoClaw, or Ottie deployment through Cloud Claw. Covers the same user-facing fields and constraints exposed in the Cloud Claw UI, using the local altllm cloud-claw-* commands. Do NOT use for post-launch lifecycle tasks like start/stop/delete/logs; use cloud-claw-manage-vm.

1k tokens
Hosted Agents V2 Py
by lingxling
×1

Build hosted agents using Azure AI Projects SDK with ImageBasedHostedAgentDefinition. Use when creating container-based agents in Azure AI Foundry.

2k tokens
Cloudflare MCP Server
by ComeOnOliver
×1

Build MCP (Model Context Protocol) servers on Cloudflare Workers with tools, resources, and prompts.

49k tokens scripts
When Chaining Agent Pipelines Use Stream Chain
by ComeOnOliver
×1

Chain agent outputs as inputs in sequential or parallel pipelines for data flow orchestration

5k tokens
Clone Audit Mrlv3nl4
by nexu-io

Audit cloned or reimplemented websites for fidelity gaps, tracking scripts, source-brand and language residue, placeholders, and risky external dependencies. Use before handoff or deployment, or when asked to review a website clone for cleanup and readiness.

2k tokens
Hermes Tweet
by wshobson

> Install and operate Hermes Tweet, a Hermes Agent plugin for X/Twitter research, timeline reading, tweet analysis, and approval-gated tweet actions. Use this skill when installing Hermes Tweet, researching X/Twitter accounts, monitoring launch signals, investigating mentions, auditing giveaways, or preparing guarded tweet actions. Use proactively when a Hermes Agent workflow needs current X/Twitter context. Requires XQUIK_API_KEY for read and action tools.

2k tokens
Arize Evaluator
by github
vendor

Handles LLM-as-judge evaluation workflows on Arize including creating/updating evaluators, running evaluations on spans or experiments, managing tasks, trigger-run operations, column mapping, and continuous monitoring. Use when the user mentions create evaluator, LLM judge, hallucination, faithfulness, correctness, relevance, run eval, score spans, score experiment, trigger-run, column mapping, continuous monitoring, or improve evaluator prompt.

10k tokens

How to use it

Copy the folder

Take majiayu000/claude-skill-registry-agent-creator from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.