Mcp Client

CLI to list and call tools on any MCP server without loading all schemas into context

Tested · Works

Test report

Verdict
Tested · Works
Score
9.2/10
Tested
Jul 21, 2026
Environment
Claude Code 2.x (agent harness)
Upstream re-checked
Aug 10, 2026 · 574491a

Cloned into a temp HOME, installed mcp+fastmcp in a venv, and ran the client against a real stdio server (@modelcontextprotocol/server-sequential-thinking). All three subcommands worked: servers returned the configured entry, tools returned the sequentialthinking tool with full JSON schema, and call executed the tool and returned its result JSON; an unknown server name returned {"type":"validation"}. Baseline: a 14-line hand-written script listed only tool names for one hardcoded stdio server; the skill's single command auto-detects 4 transports, resolves config from 4 fallback locations, and returns full schemas plus a call path. All 4 referenced files fetched HTTP 200; frontmatter parsed with name+description; no security smells (only benign "subprocess" doc mentions).

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

  • Installs cleanly 5/5
  • Triggers reliably 5/5
  • Output vs. baseline 8/10
  • Docs & honesty 5/5

What Mcp Client does

A universal command-line MCP client that connects to any configured MCP server (stdio, SSE, streamable HTTP, or FastMCP/Bearer), lists a server's tools with full parameter schemas, and executes tool calls, loading schemas on demand instead of dumping them into the context window. It triggers when the user asks to connect to an MCP server, list MCP tools, or call an MCP action (Zapier, sequential-thinking, GitHub, filesystem, etc.). Config resolves from an env var, the skill's references folder, .mcp.json, or ~/.claude.json.

How to install Mcp Client

git clone --depth 1 https://github.com/coleam00/second-brain-skills.git /tmp/mcp-client-src
mkdir -p ~/.claude/skills
cp -R /tmp/mcp-client-src/.claude/skills/mcp-client ~/.claude/skills/mcp-client
# Python deps required to run the client:
#   pip install mcp fastmcp
# Configure servers by editing ~/.claude/skills/mcp-client/references/mcp-config.json
#   (template at references/example-mcp-config.json), or set MCP_CONFIG_PATH / MCP_CONFIG,
#   or rely on an existing .mcp.json / ~/.claude.json.
# Usage:
#   python ~/.claude/skills/mcp-client/scripts/mcp_client.py servers
#   python ~/.claude/skills/mcp-client/scripts/mcp_client.py tools <server>
#   python ~/.claude/skills/mcp-client/scripts/mcp_client.py call <server> <tool> '{"arg":"value"}'

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Mcp Client

  • /mcp-client CLI to list and call tools on any MCP server without loading all schemas into context

It also activates on plain-language prompts like these:

  • Connect to the Zapier MCP server and list tools
  • Call the GitHub MCP server to list open PRs
  • Use sequential thinking via MCP to solve this

Frequently asked questions

Is the Mcp Client skill free?
Yes. The skill itself is free from coleam00/second-brain-skills. SkillProof publishes the install command and an independent test verdict at no cost.
Does Mcp Client work with Claude Code?
We tested it with Claude Code 2.x (agent harness) on Jul 21, 2026. Verdict: Tested · Works. Cloned into a temp HOME, installed mcp+fastmcp in a venv, and ran the client against a real stdio server (@modelcontextprotocol/server-sequential-thinking). All three subcommands worked: servers returned the configured entry, tools returned the sequentialthinking tool with full JSON schema, and call executed the tool and returned its result JSON; an unknown server name returned {"type":"validation"}. Baseline: a 14-line hand-written script listed only tool names for one hardcoded stdio server; the skill's single command auto-detects 4 transports, resolves config from 4 fallback locations, and returns full schemas plus a call path. All 4 referenced files fetched HTTP 200; frontmatter parsed with name+description; no security smells (only benign "subprocess" doc mentions).
What is the Mcp Client SkillProof Score?
9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
How do I install Mcp Client?
Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Mcp Client with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.