Lightweight, replayable PoC for Sovereign Epistemic Commons: synthetic multi-agent memory-governance experiments on contamination, typed lanes, provenance-depth discount, and mixed exit/fork in observable-only agentic systems.
ai provenance shared-memory multi-agent-systems contamination replayability llm-evaluation agentic-systems memory-governance knowledge-governance epistemic-commons observable-only sovereign-epistemic-commons synthetic-benchmark contradiction-handling typed-lanes fork-recovery cooldown-policy
-
Updated
Mar 17, 2026 - Python