The sources of valid cognition (Pramāṇa) are direct perception, inference, and reliable testimony.

For a data packet to be flagged as Pramāṇa (Valid Cognition), it must originate from one of three validated input channels:

Pratyakṣa (Direct Perception): Data received through the sensory interface. This is a direct input channel providing immediate, minimally processed signal.

Anumāṇa (Logical Inference): Data derived through structured reasoning. If A = B and B = C, the system infers A = C. It is a computed output based on previously validated inputs.

Āgama (Reliable Testimony): Data obtained from trusted and verified external sources. This includes knowledge transmitted through reliable systems, not limited to texts or authorities. It functions like a validated external input stream.

प्रत्यक्षानुमानागमाः प्रमाणानि

pratyakṣa-anumāna-āgamāḥ pramāṇāni

— Yoga Sūtra 1.07

正知来源

正知（Pramāṇa）的来源有三种：现量（直接感知）、比量（逻辑推论）与圣言量（可靠见证）。

在 Mind OS 中，一条数据若要被标记为 Pramāṇa（正知），必须来自以下三种已验证的输入通道之一：

Pratyakṣa（现量/直接感知）： 通过感官接口接收的数据。这是一种直接输入通道，提供即时且最少加工的信号。

Anumāṇa（比量/逻辑推论）： 通过结构化推理得出的数据。如果 A = B 且 B = C，则系统推导出 A = C。它是基于已验证输入的计算结果。

Āgama（圣言量/可靠见证）： 来自受信任且经过验证的外部来源的数据。这并不局限于经典或权威，而是任何可靠的知识传递系统。

प्रत्यक्षानुमानागमाः प्रमाणानि

pratyakṣa-anumāna-āgamāḥ pramāṇāni

——《瑜伽经》Yoga Sūtra 1.07