2
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

MCPクライアント対応機能のマトリクス

Posted at

概要

MCPホストを選択する時、公式サイトの対応機能を参考にする。しかし、フィルター機能しか用意されておらず、比較が困難である。
今回は、対応機能のマトリクス表とそのマトリクス表を取得するコードを作成した。

image.png

Client details マトリクス(2026年1月25日時点)

Name Discovery Elicitation Instructions Prompts Resources Resources (partial) Roots Sampling Sampling (partial) Tasks Tools
5ire × × × × × × × × × ×
AgentAI × × × × × × × × × ×
AgenticFlow × × × × × × ×
AIQL TUUI × × × × ×
Amazon Q CLI × × × × × × × × ×
Amazon Q IDE × × × × × × × × × ×
Amp × × × × × × ×
Apify MCP Tester × × × × × × × × ×
Augment Code × × × × × × × × × ×
Avatar Shell × × × × × × × × ×
BeeAI Framework × × × × × × × × × ×
BoltAI × × × × × × × × × ×
Call Chirp × × × × × × × × ×
Chatbox × × × × × × × × × ×
ChatFrame × × × × × × × × × ×
ChatGPT × × × × × × × × × ×
ChatWise × × × × × × × × × ×
Chorus × × × × × × × × × ×
Claude Code × × × × ×
Claude Desktop App × × × × × × × ×
Claude.ai × × × × × × × ×
Cline × × × × × × × ×
CodeGPT × × × × × × × × × ×
Codex × × × × × × × ×
Continue × × × × × × × ×
Copilot-MCP × × × × × × × × ×
Cursor × × × × × × ×
Daydreams × × × × × × × ×
ECA - Editor Code Assistant × × × × × × ×
Emacs Mcp × × × × × × × × × ×
fast-agent × × ×
Firebender × × × × × × × × × ×
FlowDown × × × × × × × × × ×
FLUJO × × × × × × × × × ×
Gemini CLI × × × × × × × ×
GenAIScript × × × × × × × × ×
Genkit × × × × × × × ×
GitHub Copilot coding agent × × × × × × × × × ×
Glama × × ×
goose × × × × ×
gptme × × × × × × × × × ×
HyperAgent × × × × × × × × × ×
Jenova × × × × × × × × ×
JetBrains AI Assistant × × × × × × × × × ×
JetBrains Junie × × × × × × × × × ×
Kilo Code × × × × × × × ×
Klavis AI Slack/Discord/Web × × × × × × × × ×
Langdock × × × × × × × × × ×
Langflow × × × × × × × × × ×
LibreChat × × × × × × × × ×
LM Studio × × × × × × × × × ×
LM-Kit.NET × × × × × × × × × ×
Lutra × × × × × × × ×
MCP Bundler for MacOS × × × × × × × ×
MCPBundles × × × × × × ×
mcp-agent × × × × ×
mcp-client-chatbot × × × × × × × × × ×
mcp-use × × × × ×
mcpc × × × × × ×
MCPHub × × × × × × × ×
MCPJam × × × × × × ×
MCPOmni-Connect × × × × × × ×
Memex × × × × × × × ×
Memgraph Lab × × × × × × ×
Microsoft Copilot Studio × × × × × × × ×
MindPal × × × × × × × × × ×
Mistral AI: Le Chat × × × × × × × × × ×
modelcontextchat.com × × × × × × × × × ×
MooPoint × × × × × × × × ×
Msty Studio × × × × × × × × × ×
Needle × × × × × × ×
NVIDIA Agent Intelligence (AIQ) toolkit × × × × × × × × × ×
opencode × × × × × × × ×
OpenSumi × × × × × × × × × ×
oterm × × × × × × × ×
Postman × × × × ×
RecurseChat × × × × × × × × × ×
Replit × × × × × × × × × ×
Roo Code × × × × × × × × ×
rtrvr.ai × × × × × × × × × ×
Shortwave × × × × × × × × × ×
Simtheory × × × × × × ×
Slack MCP Client × × × × × × × × × ×
Smithery Playground × × × × × × × ×
SpinAI × × × × × × × × × ×
Superinterface × × × × × × × × × ×
Superjoin × × × × × × × × × ×
Swarms × × × × × × × × ×
systemprompt × × × × × × ×
Tambo × × × × × ×
Tencent CloudBase AI DevKit × × × × × × × × × ×
TheiaAI/TheiaIDE × × × × × × × × × ×
Tome × × × × × × × × × ×
TypingMind App × × × × × × × × × ×
v0 × × × × × × × × × ×
VS Code GitHub Copilot × × ×
VT Code × × × ×
Warp × × × × × × × ×
WhatsMCP × × × × × × × × × ×
Windsurf Editor × × × × × × × × ×
Witsy × × × × × × × × × ×
Zed × × × × × × × × ×
Zencoder × × × × × × × × × ×

今回のコードでできること

以下のサイトで「Copy page」をクリックして「mcp_client.md」に張り付けて保存し、

image.png

「mcp_client_support_matrix.py」を実行すればマトリクス表が得られる。
image.png

ソースコード(mcp_client_support_matrix.py)

import re
from pathlib import Path

# =========================
# 設定
# =========================
MD_PATH = Path("mcp_client.md")

# =========================
# McpClient 抽出
# =========================
text = MD_PATH.read_text(encoding="utf-8")

client_pattern = re.compile(r"<McpClient\s+([^>]+)>", re.IGNORECASE)
attr_pattern = re.compile(r'(\w+)="([^"]*)"')

clients = []
all_supports = set()

for m in client_pattern.finditer(text):
    attrs = dict(attr_pattern.findall(m.group(1)))
    if "name" in attrs and "supports" in attrs:
        supports = [s.strip() for s in attrs["supports"].split(",")]
        clients.append({
            "name": attrs["name"],
            "supports": set(supports),
        })
        all_supports.update(supports)

support_types = sorted(all_supports)

# =========================
# Markdown テーブル生成
# =========================
header = ["Name"] + support_types
rows = []

for c in clients:
    row = [c["name"]]
    for s in support_types:
        row.append("" if s in c["supports"] else "×")
    rows.append(row)

# =========================
# Markdown 出力
# =========================
def md_row(cols):
    return "| " + " | ".join(cols) + " |"

print(md_row(header))
print(md_row(["---"] * len(header)))

for row in rows:
    print(md_row(row))
2
1
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
2
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?