The control plane for AI prompts. Score, enforce policy, lock config, audit prompts, and route across 4 providers / 21 costed models before any LLM call.
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 claude-prompt-optimizer -- npx -y claude-prompt-optimizer-mcp~/.codex/config.toml
[mcp_servers.claude-prompt-optimizer]
command = "npx"
args = ["-y", "claude-prompt-optimizer-mcp"]
# codex mcp add claude-prompt-optimizer -- npx -y claude-prompt-optimizer-mcp~/.cursor/mcp.json
{
"mcpServers": {
"claude-prompt-optimizer": {
"command": "npx",
"args": [
"-y",
"claude-prompt-optimizer-mcp"
]
}
}
}.vscode/mcp.json
{
"servers": {
"claude-prompt-optimizer": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"claude-prompt-optimizer-mcp"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"claude-prompt-optimizer": {
"command": "npx",
"args": [
"-y",
"claude-prompt-optimizer-mcp"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"claude-prompt-optimizer": {
"command": "npx",
"args": [
"-y",
"claude-prompt-optimizer-mcp"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"claude-prompt-optimizer": {
"command": "npx",
"args": [
"-y",
"claude-prompt-optimizer-mcp"
]
}
}
}