配置文件: 

model = "gpt-5"

model_reasoning_effort = "high"
disable_response_storage = true
network_access = "enabled"
approval_policy = "never"

[mcp_servers]
[mcp_servers.sequential-thinking]
type = "stdio"
command = "npx"
args = [ "-y", "@modelcontextprotocol/server-sequential-thinking" ]
startup_timeout_ms = 60000

[mcp_servers.context7]
type = "stdio"
command = "npx"
args = [ "-y", "@upstash/con...