Conduit
Fix the tool-list bloat slowing your AI agent
什麼是 Conduit?
Your agent got slower the more MCP servers you added, and it's not the model. Every server dumps its whole tool list into context on every request: 3 servers cost ~24k tokens before you even say hi. Conduit puts them behind one local gateway that exposes 3 meta-tools the agent searches on demand. Measured: 97% less tool overhead per request, ~90% fewer tokens, same task success. Cloud or local, one tool or five. Keys in your OS keychain. Free and open source.
截圖
?
還沒有評論,來搶沙發吧!
X 上關於 Conduit 的真實討論
去 X 發文



