
Agents
- Updated May 7, 2026
- CHENyiru3/AI-Skills-Collections
agents is an agent skill that curates coding-agent and MCP skill paths—Claude Code, Codex, Hermes, OpenCode, guidelines, and MCP builder—usable whenever a solo builder needs to pick or configure agent tooling
About
The agents skill is a Prism-facing bundle from AI-Skills-Collections that groups the skills-market agents folder: dedicated subtrees for Claude Code, Codex, Hermes Agent, OpenCode, Zed, Karpathy guidelines, and an MCP developer path. developers use it as a map and install anchor when they are choosing or rotating coding agents, wiring MCP servers, or aligning behavior with published agent guidelines rather than guessing which repo folder to open. Because agent choice affects every phase—from Idea research with Codex to Grow automation with MCP—it is journey-wide in scope even though the catalog shelf sits in Build agent-tooling. Invoke it when onboarding a new machine, comparing Hermes versus Claude Code workflows, or drafting an MCP integration plan. It does not replace deep SKILL.md content inside each child skill; it routes you to them.
- Aggregates Claude Code, Codex, Hermes Agent, and OpenCode skill entry points
- Includes MCP builder material under agents/mcp for extending tool access
- Bundles Karpathy guidelines and Zed-oriented agent guidance for consistency
- Market layout with strict:false so you can mix agent sub-skills flexibly
- Single description surface for comparing agent stacks before you commit to one CLI
Agents by the numbers
- Data as of Jul 7, 2026 (Skillselion catalog sync)
/plugin marketplace add CHENyiru3/AI-Skills-Collections/plugin install agents@ai-skillsAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Last updated | May 7, 2026 |
|---|---|
| Repository | CHENyiru3/AI-Skills-Collections ↗ |
What it does
Install the agents market bundle when you need curated paths into Claude Code, Codex, Hermes, OpenCode, MCP developer, Zed, and Karpathy-style guidelines in one place.
Who is it for?
Best when you're exploring or standardizing among Claude Code, Codex, Hermes, OpenCode, and MCP and want one catalog slug that lists all agent-related skills.
Skip if: Skip if you already pinned one agent skill and never switch tools, or teams needing SAP/enterprise vertical skills with no agent onboarding need.
When should I use this skill?
When configuring, comparing, or extending coding agents—Claude Code, Codex, Hermes, OpenCode, MCP, or Zed—or applying bundled agent guidelines.
What you get
You land on the right child skill—Claude Code, Codex, MCP builder, or guidelines—with a consistent market entry instead of searching the monorepo blindly.
- Resolved path to the correct nested agent or MCP skill
- Aligned agent guideline references for subsequent tasks
By the numbers
- Seven named child skill paths in the market manifest (Claude Code, Codex, Karpathy guidelines, Hermes, MCP builder, Open
Files
Coding agents — Claude Code, Codex, Hermes, OpenCode, MCP, guidelines
agents
Source: ./skills-market/agents
{ "name": "agents", "skills": [ "./agents/claude-code", "./agents/codex", "./agents/guidelines/karpathy-guidelines", "./agents/hermes-agent", "./agents/mcp/mcp-builder", "./agents/opencode", "./agents/zed" ], "source": "./skills-market/agents", "strict": false, "description": "Coding agents — Claude Code, Codex, Hermes, OpenCode, MCP, guidelines" }
Related skills
How it compares
Meta skill-market index for agent stacks—not a deploy script, MCP server binary, or single-vendor API integration.
FAQ
Who is agents for?
agents is for developers and small teams who use or compare Claude Code, Codex, Hermes, OpenCode, Zed, and MCP patterns and want a single skills-market entry for those paths.
When should I use agents?
Use it at journey start when picking an agent CLI, during Build when wiring MCP, before Ship when aligning review bots with Karpathy guidelines, and anytime you switch agent tools mid-project.
Is agents safe to install?
The bundle points at multiple child skills with different permissions; check the Security Audits panel on this page and review each nested skill before enabling network or shell tools.