native-mcp

Connect MCP servers via stdio or HTTP and register their tools automatically.

Updated Jul 10, 2026
One-click install
npx skills add https://github.com/AvaTar-ArTs/.Agent-skills --skill native-mcp-avatar-arts
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: native-mcp
Source: https://github.com/AvaTar-ArTs/.Agent-skills/tree/main/skills/mcp/native-mcp
Command: npx skills add https://github.com/AvaTar-ArTs/.Agent-skills --skill native-mcp-avatar-arts

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mcp.

What problem does it solve? Manually wiring external tools into an AI agent requires custom bridge code and repetitive integration work. This Skill configures Hermes Agent's built-in MCP client so tools from any MCP server are discovered at startup and exposed as first-class callable tools. ## Core Features & Use Cases - Automatic Tool Discovery: Connects to configured MCP servers at startup, calls list_tools(), and registers each tool with an mcp_{server}{tool} naming prefix. - Dual Transport Support: Runs local stdio servers via npx or uvx commands, and connects to remote servers over HTTP/StreamableHTTP with custom headers. - Security Controls: Filters environment variables passed to subprocesses, redacts credential patterns from error messages, and supports per-server sampling limits. - Use Case: Add a GitHub MCP server with a personal access token to your config, restart the agent, and immediately ask it to list issues or create pull requests using the auto-registered mcp_github* tools. ## Quick Start Add an mcp_servers entry with a command and args to ~/.hermes/config.yaml, then restart the agent and ask it to use the newly available tools.

Frequently Asked Questions about native-mcp

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I connect an MCP server to my AI agent?

Add an entry under the mcp_servers key in ~/.hermes/config.yaml with either a command and args for stdio servers or a url for HTTP servers. Restart the agent and its tools are discovered and registered automatically with an mcp_{server}_{tool} prefix.

What is the difference between stdio and HTTP MCP transports?

Stdio transport launches the MCP server as a local subprocess via commands like npx or uvx and communicates over stdin/stdout. HTTP transport connects to a remote server URL and supports custom headers such as Authorization tokens.

Does the MCP client pass my environment variables to servers?

No, stdio subprocesses receive only a filtered baseline set like PATH, HOME, and XDG_* variables. API keys and secrets must be explicitly declared in the server's env config block to be passed through.

Why are my MCP tools not appearing after configuration?

Common causes include a missing mcp Python package, incorrect YAML indentation, the server key not being under mcp_servers, or the command binary not being on PATH. Check startup logs and look for tool names with the mcp_{server}_{tool} pattern.

Can MCP servers request LLM completions during tool execution?

Yes, sampling via sampling/createMessage is enabled by default and configurable per server. You can set model overrides, token caps, rate limits, allowed models, and tool round limits, or disable it entirely for untrusted servers.