Cognitive Architectures for Language Agents
A framework for reasoning about an agent's memory, actions, and decision loop
Topic
Artificial Intelligence
7 notes
August 2026
The Phoenix Architecture
Regenerating disposable implementations from durable intent and evaluations
Cognitive Surrender and Tri-System Theory
How AI can support, displace, or bypass human deliberation
MCP and Agent Skills
Connecting agents to capabilities and teaching them how to use those capabilities
Embeddings
Learned vector representations of meaning and similarity
Retrieval-Augmented Generation
Grounding language-model answers in retrieved source material
Vectors
How vectors represent data and make similarity search possible