Python API

YantrikDB

The main database class.

Constructor

db = YantrikDB(db_path="memory.db", embedding_dim=384)

Parameter	Type	Default	Description
db_path	str	"memory.db"	Path to the database file
embedding_dim	int	384	Embedding vector dimensions

Core Methods

record(text, **kwargs)

Store a memory.

db.record(
    "User likes Python",
    importance=0.7,      # 0.0-1.0
    valence=0.3,         # -1.0 to 1.0 (emotional)
    memory_type="semantic",  # episodic, semantic, procedural
    domain="work",
    source="user",
)

recall(query, top_k=10, **kwargs)

Retrieve memories using relevance-conditioned scoring.

results = db.recall(
    "What programming languages?",
    top_k=5,
    memory_type="semantic",  # optional filter
    namespace="default",     # optional filter
)

Returns a list of dicts with text, score, rid, importance, valence, etc.

relate(src, dst, rel_type, weight=1.0)

Create a relationship in the cognitive graph.

db.relate("Pranab", "Walmart", "works_at", weight=0.9)

think(**kwargs)

Run autonomous cognition: consolidation, conflict detection, pattern mining, triggers.

result = db.think(
    importance_threshold=0.5,
    run_consolidation=True,
    run_conflict_scan=True,
    run_pattern_mining=True,
)

Returns a dict with triggers, consolidation_count, conflicts_found, patterns_new, duration_ms.

forget(rid)

Tombstone a memory by record ID.

db.forget("019d0a1b-...")

stats()

Get database statistics.

s = db.stats()
# s['active_memories'], s['edges'], s['entities'], etc.

Graph Methods

entities()

List all entities in the cognitive graph.

edges(entity=None)

List edges, optionally filtered by entity.

conflicts()

List all detected contradictions.

patterns()

List all discovered behavioral patterns.

Embedding

set_embedder(fn)

Set the embedding function.

from sentence_transformers import SentenceTransformer
model = SentenceTransformer("all-MiniLM-L6-v2")
db.set_embedder(lambda text: model.encode(text).tolist())