https://multi-agent-llm-teams.github.io/ 2026-10-31 monthly 1.0 https://multi-agent-llm-teams.github.io/static/images/hero_buffering.png Artifact-mediated buffering: low agreeableness shifts communication identically in every domain, but only unstructured tasks see outcomes degrade; structured code output buffers the effect https://multi-agent-llm-teams.github.io/static/images/og.jpg When Does Personality Composition Matter for Multi-Agent LLM Teams? (COLM 2026)