Adaptation of the Sequential Thinking server that breaks problems into steps and recommends which MCP tools to use at each stage.
接入配置
选择客户端，复制配置，再按项目文档补齐环境变量。
claude mcp add sequential-thinking-tools --env MAX_HISTORY_SIZE='<MAX_HISTORY_SIZE>' -- npx -y mcp-sequentialthinking-tools~/.codex/config.toml
[mcp_servers.sequential-thinking-tools]
command = "npx"
args = ["-y", "mcp-sequentialthinking-tools"]
env = { MAX_HISTORY_SIZE = "<MAX_HISTORY_SIZE>" }
# codex mcp add sequential-thinking-tools --env MAX_HISTORY_SIZE='<MAX_HISTORY_SIZE>' -- npx -y mcp-sequentialthinking-tools~/.cursor/mcp.json
{
"mcpServers": {
"sequential-thinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
],
"env": {
"MAX_HISTORY_SIZE": "<MAX_HISTORY_SIZE>"
}
}
}
}.vscode/mcp.json
{
"servers": {
"sequential-thinking-tools": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
],
"env": {
"MAX_HISTORY_SIZE": "<MAX_HISTORY_SIZE>"
}
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"sequential-thinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
],
"env": {
"MAX_HISTORY_SIZE": "<MAX_HISTORY_SIZE>"
}
}
}
}claude_desktop_config.json
{
"mcpServers": {
"sequential-thinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
],
"env": {
"MAX_HISTORY_SIZE": "<MAX_HISTORY_SIZE>"
}
}
}
}.trae/mcp.json
{
"mcpServers": {
"sequential-thinking-tools": {
"command": "npx",
"args": [
"-y",
"mcp-sequentialthinking-tools"
],
"env": {
"MAX_HISTORY_SIZE": "<MAX_HISTORY_SIZE>"
}
}
}
}需要的环境变量
MAX_HISTORY_SIZE
把占位值换成你自己的密钥；不要把密钥提交到代码仓库。