Enforce the allowed versions of code that can perform an action

未關閉

#3 建立於 2025年10月20日 0 則留言 0 個 reaction 已指派 0 人在 GitHub 檢視

描述

We need to improve the security posture of this process beyond allowing selected clients to guarantee information integrity.

The ideas is that we can have guardrails to avoid an actor to modify the code and process the AI task with bad intent.

這個儲存庫沒有索引到貢獻指南

這個 Issue 還沒有評估資料。