Skip to main content
Kremis tracks the maturity of the graph through developmental stages based on the count of stable edges (edges with weight ≥ 10).
Stages are informational only — they do not gate functionality.

Stage Progression

StageNameStable Edge Threshold
S0Signal Segmentation0
S1Pattern Crystallization100
S2Causal Chaining1,000
S3Recursive Optimization5,000

What Each Stage Represents

1

S0 — Signal Segmentation

Initial stage. The graph is building basic structure from incoming signals. Few or no stable patterns yet.
2

S1 — Pattern Crystallization

Patterns are emerging from repeated signals. At least 100 edges have been reinforced to stability (weight ≥ 10).
3

S2 — Causal Chaining

Strong interconnections are forming. The graph contains meaningful chains of stable relationships.
4

S3 — Recursive Optimization

The graph is well-connected and stable. Over 5,000 edges have been reinforced through repeated observation.

Querying the Stage

HTTP API

curl http://localhost:8080/stage
{
  "stage": "S1",
  "name": "Pattern Crystallization",
  "progress_percent": 45,
  "stable_edges_needed": 100,
  "stable_edges_current": 45
}

CLI

kremis stage --detailed
Last modified on February 19, 2026