CodexSaver: Cut AI Costs by Offloading Low-Risk Tasks to DeepSeek

Saving to the MCP tools collection: fendouai/CodexSavergithub.com/fendouai/CodexSaver

It offloads low-risk tasks from Codex to DeepSeek, leaving only decision-making to the expensive models. In tests on five tasks, the average savings were 48%, with latency of about 6 seconds.