AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
翻譯準備中
正在載入完整 README。 以下為已儲存的摘錄。
A single thread may seem insignificant, but when many threads are precisely interwoven, they can be woven into any shape.
Loom is a harness-first, memory-native, self-directing agent framework for anyone who cares not just about what an agent can do — but about what it did, who approved it, and what happens when it goes wrong.
Most agent frameworks treat tools as functions: define → call → done. That works in demos. It falls apart the moment you need audit trails, real authorization logic, OS-level sandboxing, long-term memory, or reliable autonomous operation.檢視完整 README ↗目前無法載入完整 README。請透過原文連結閱讀。 以下為已儲存的摘錄。
A single thread may seem insignificant, but when many threads are precisely interwoven, they can be woven into any shape.
Loom is a harness-first, memory-native, self-directing agent framework for anyone who cares not just about what an agent can do — but about what it did, who approved it, and what happens when it goes wrong.
Most agent frameworks treat tools as functions: define → call → done. That works in demos. It falls apart the moment you need audit trails, real authorization logic, OS-level sandboxing, long-term memory, or reliable autonomous operation.檢視完整 README ↗