EAG schema primitives for building epistemic context systems.
pip install engrammic-primitivesfrom primitives.schema import PersistenceLayer, MemoryLabel, KnowledgeLabel
from primitives.eag import combined_confidence, should_promote_r1
# Check persistence layer for a label
layer = PersistenceLayer.MEMORY
assert MemoryLabel.UTTERANCE.layer == layer
# Combine confidence scores
conf = combined_confidence(base=0.7, corroboration=0.85)
# Check if a claim should promote to Fact
decision = should_promote_r1(confidence=0.8, corroboration_count=3)Building your own EAG-compatible system or extending Engrammic.
For using Engrammic directly, see:
- engrammic-mcp - hosted service
- engrammic-engine - local engine
- EAG Paradigm - the four-layer cognitive architecture
| Module | Purpose |
|---|---|
primitives.schema |
Node and edge type definitions |
primitives.eag |
EAG-specific implementations |
primitives.protocols |
Storage and lifecycle interfaces |
primitives.scoring |
Decay and freshness formulas |
Apache 2.0