SKILL·29832E

mcp-cli

obra
更新于 22 days ago
61 次查看
432
36
432
在 GitHub 上查看
开发mcp

关于

This skill enables on-demand usage of MCP servers through the mcp CLI tool, allowing developers to dynamically discover and invoke tools without permanent integrations. It's ideal for exploring server capabilities, making one-off calls, or testing MCP functionality while avoiding context pollution. The skill requires the mcp CLI to be installed and available in the system PATH.

快速安装

Claude Code

推荐
主要方式
npx skills add obra/superpowers-lab -a claude-code
插件命令备选方式
/plugin add https://github.com/obra/superpowers-lab
Git 克隆备选方式
git clone https://github.com/obra/superpowers-lab.git ~/.claude/skills/mcp-cli

在 Claude Code 中复制并粘贴此命令以安装该技能

技能文档

MCP CLI: On-Demand MCP Server Usage

Use the mcp CLI tool to dynamically discover and invoke MCP server capabilities without pre-configuring them as permanent integrations.

When to Use This Skill

Use this skill when you need to:

  • Explore an MCP server's capabilities before deciding to use it
  • Make one-off calls to an MCP server without permanent integration
  • Access MCP functionality without polluting the context window
  • Test or debug MCP servers
  • Use MCP servers that aren't pre-configured

Prerequisites

The mcp CLI must be installed at ~/.local/bin/mcp. If not present:

# Clone and build
cd /tmp && git clone --depth 1 https://github.com/f/mcptools.git
cd mcptools && CGO_ENABLED=0 go build -o ~/.local/bin/mcp ./cmd/mcptools

Always ensure PATH includes the binary:

export PATH="$HOME/.local/bin:$PATH"

Discovery Workflow

Step 1: Discover Available Tools

mcp tools <server-command>

Examples:

# Filesystem server
mcp tools npx -y @modelcontextprotocol/server-filesystem /path/to/allow

# Memory/knowledge graph server
mcp tools npx -y @modelcontextprotocol/server-memory

# GitHub server (requires token)
mcp tools docker run -i --rm -e GITHUB_PERSONAL_ACCESS_TOKEN ghcr.io/github/github-mcp-server

# HTTP-based server
mcp tools https://example.com/mcp

Step 2: Discover Resources (if supported)

mcp resources <server-command>

Resources are data sources the server exposes (files, database entries, etc.).

Step 3: Discover Prompts (if supported)

mcp prompts <server-command>

Prompts are pre-defined prompt templates the server provides.

Step 4: Get Detailed Info (JSON format)

# For full schema details including parameter types
mcp tools --format json <server-command>
mcp tools --format pretty <server-command>

Making Tool Calls

Basic Syntax

mcp call <tool_name> --params '<json>' <server-command>

Examples

Read a file:

mcp call read_file --params '{"path": "/tmp/example.txt"}' \
  npx -y @modelcontextprotocol/server-filesystem /tmp

Write a file:

mcp call write_file --params '{"path": "/tmp/test.txt", "content": "Hello world"}' \
  npx -y @modelcontextprotocol/server-filesystem /tmp

List directory:

mcp call list_directory --params '{"path": "/tmp"}' \
  npx -y @modelcontextprotocol/server-filesystem /tmp

Create entities (memory server):

mcp call create_entities --params '{"entities": [{"name": "Project", "entityType": "Software", "observations": ["Uses TypeScript"]}]}' \
  npx -y @modelcontextprotocol/server-memory

Search (memory server):

mcp call search_nodes --params '{"query": "TypeScript"}' \
  npx -y @modelcontextprotocol/server-memory

Complex Parameters

For nested objects and arrays, ensure valid JSON:

mcp call edit_file --params '{
  "path": "/tmp/file.txt",
  "edits": [
    {"oldText": "foo", "newText": "bar"},
    {"oldText": "baz", "newText": "qux"}
  ]
}' npx -y @modelcontextprotocol/server-filesystem /tmp

Output Formats

# Table (default, human-readable)
mcp call <tool> --params '{}' <server>

# JSON (for parsing)
mcp call <tool> --params '{}' -f json <server>

# Pretty JSON (readable JSON)
mcp call <tool> --params '{}' -f pretty <server>

Reading Resources

# List available resources
mcp resources <server-command>

# Read a specific resource
mcp read-resource <resource-uri> <server-command>

# Alternative syntax
mcp call resource:<resource-uri> <server-command>

Using Prompts

# List available prompts
mcp prompts <server-command>

# Get a prompt (may require arguments)
mcp get-prompt <prompt-name> <server-command>

# With parameters
mcp get-prompt <prompt-name> --params '{"arg": "value"}' <server-command>

Server Aliases (for repeated use)

If using a server frequently during a session:

# Create alias
mcp alias add fs npx -y @modelcontextprotocol/server-filesystem /home/user

# Use alias
mcp tools fs
mcp call read_file --params '{"path": "README.md"}' fs

# List aliases
mcp alias list

# Remove when done
mcp alias remove fs

Aliases are stored in ~/.mcpt/aliases.json.

Authentication

HTTP Basic Auth

mcp tools --auth-user "username:password" https://api.example.com/mcp

Bearer Token

mcp tools --auth-header "Bearer your-token-here" https://api.example.com/mcp

Environment Variables (for Docker-based servers)

mcp tools docker run -i --rm \
  -e GITHUB_PERSONAL_ACCESS_TOKEN="$GITHUB_TOKEN" \
  ghcr.io/github/github-mcp-server

Transport Types

Stdio (default for npx/node commands)

mcp tools npx -y @modelcontextprotocol/server-filesystem /tmp

HTTP (auto-detected for http/https URLs)

mcp tools https://example.com/mcp

SSE (Server-Sent Events)

mcp tools http://localhost:3001/sse
# Or explicitly:
mcp tools --transport sse http://localhost:3001

Common MCP Servers

Filesystem

# Allow access to specific directory
mcp tools npx -y @modelcontextprotocol/server-filesystem /path/to/allow

Memory (Knowledge Graph)

mcp tools npx -y @modelcontextprotocol/server-memory

GitHub

export GITHUB_PERSONAL_ACCESS_TOKEN="your-token"
mcp tools docker run -i --rm -e GITHUB_PERSONAL_ACCESS_TOKEN ghcr.io/github/github-mcp-server

Brave Search

export BRAVE_API_KEY="your-key"
mcp tools npx -y @anthropic/mcp-server-brave-search

Puppeteer (Browser Automation)

mcp tools npx -y @anthropic/mcp-server-puppeteer

Best Practices

1. Always Discover First

Before calling tools, run mcp tools to understand what's available and the exact parameter schema.

2. Use JSON Format for Parsing

When you need to process results programmatically:

mcp call <tool> --params '{}' -f json <server> | jq '.field'

3. Validate Parameters

The table output shows parameter signatures. Match them exactly:

  • param:str = string
  • param:num = number
  • param:bool = boolean
  • param:str[] = array of strings
  • [param:str] = optional parameter

4. Handle Errors Gracefully

Tool calls may fail. Check exit codes and stderr:

if ! result=$(mcp call tool --params '{}' server 2>&1); then
  echo "Error: $result"
fi

5. Use Aliases for Multi-Step Operations

If making several calls to the same server:

mcp alias add tmp-server npx -y @modelcontextprotocol/server-filesystem /tmp
mcp call list_directory --params '{"path": "/tmp"}' tmp-server
mcp call read_file --params '{"path": "/tmp/file.txt"}' tmp-server
mcp alias remove tmp-server

6. Restrict Capabilities with Guard

For safety, limit what tools are accessible:

# Only allow read operations
mcp guard --allow 'tools:read_*,list_*' --deny 'tools:write_*,delete_*' \
  npx -y @modelcontextprotocol/server-filesystem /home

Debugging

View Server Logs

mcp tools --server-logs <server-command>

Check Alias Configuration

cat ~/.mcpt/aliases.json

Verbose Output

Use --format pretty for detailed JSON output to debug parameter issues.

Quick Reference

ActionCommand
List toolsmcp tools <server>
List resourcesmcp resources <server>
List promptsmcp prompts <server>
Call toolmcp call <tool> --params '<json>' <server>
Read resourcemcp read-resource <uri> <server>
Get promptmcp get-prompt <name> <server>
Add aliasmcp alias add <name> <server-command>
Remove aliasmcp alias remove <name>
JSON outputAdd -f json or -f pretty

Example: Complete Workflow

# 1. Discover what's available
mcp tools npx -y @modelcontextprotocol/server-filesystem /home/user/project

# 2. Check for resources
mcp resources npx -y @modelcontextprotocol/server-filesystem /home/user/project

# 3. Create alias for convenience
mcp alias add proj npx -y @modelcontextprotocol/server-filesystem /home/user/project

# 4. Explore directory structure
mcp call directory_tree --params '{"path": "/home/user/project"}' proj

# 5. Read specific files
mcp call read_file --params '{"path": "/home/user/project/README.md"}' proj

# 6. Search for patterns
mcp call search_files --params '{"path": "/home/user/project", "pattern": "**/*.ts"}' proj

# 7. Clean up alias
mcp alias remove proj

Troubleshooting

"command not found: mcp"

Ensure PATH is set: export PATH="$HOME/.local/bin:$PATH"

JSON parse errors

  • Escape special characters properly
  • Avoid shell expansion issues by using single quotes around JSON
  • For complex JSON, write to a temp file and use --params "$(cat params.json)"

Server timeout

Some servers take time to start. The mcp CLI waits for initialization automatically.

Permission denied

For filesystem server, ensure the allowed directory path is correct and accessible.

GitHub 仓库

obra/superpowers-lab
路径: skills/mcp-cli
0
FAQ

常见问题

什么是 mcp-cli Skill?

mcp-cli 是一个 Claude Skill,作者为 obra。Skill 将 Claude 按需加载的说明和资源打包,让 Claude 无需额外提示即可执行与 mcp-cli 相关的任务。

如何安装 mcp-cli?

使用本页的安装命令:将 mcp-cli 作为插件添加到 Claude Code,或将其仓库克隆到 skills 目录,然后重启 Claude 以加载该 Skill。

mcp-cli 属于哪个分类?

mcp-cli 属于开发分类。

mcp-cli 可以免费使用吗?

可以。mcp-cli 已收录在 AIMCP,可免费安装。

相关推荐技能

qmd
开发

这是一个本地搜索和索引的CLI工具,支持BM25、向量搜索和重排序功能。开发者可以用它快速索引本地文件(如Markdown文档)并进行混合搜索,特别适合代码库或文档的本地检索。它还提供MCP模式,能轻松集成到Claude开发环境中使用。

查看技能
subagent-driven-development
开发

该Skill用于在当前会话中执行包含独立任务的实施计划,它会为每个任务分派一个全新的子代理并在任务间进行代码审查。这种"全新子代理+任务间审查"的模式既能保障代码质量,又能实现快速迭代。适合需要在当前会话中连续执行独立任务,并希望在每个任务后都有质量把关的开发场景。

查看技能
mcporter
开发

mcporter Skill 让开发者能在Claude中直接管理和调用MCP服务器。它支持列出可用服务器、调用工具、处理OAuth认证以及管理服务器守护进程。开发者可以通过命令行式交互快速执行`mcporter list`查看服务器,或使用`mcporter call`直接调用工具,简化了MCP工作流程。

查看技能
adk-deployment-specialist
开发

这是一个用于部署和编排Google Vertex AI ADK智能体的Claude Skill,专为构建生产级多智能体系统而设计。它支持通过A2A协议进行智能体通信,提供代码执行沙箱和记忆库功能,并能处理智能体发现与任务提交。当开发者需要部署ADK智能体或编排多智能体协作时,可使用此Skill来简化Vertex AI Agent Engine的部署流程。

查看技能