What problem does it solve?
This Skill prevents duplicate and conflicting records by deterministically resolving incoming entity records to the correct canonical entity_id using evidence-based, field-level matching.
Core Features & Use Cases
- Deterministic identity resolution: Produces the same canonical entity_id for the same real-world entity even across concurrent writes and multi-agent access.
- Evidence-based fuzzy matching: Uses blocking plus field-level scoring (including nickname normalization and E.164 phone formatting) and explains results with confidence and per-field evidence.
- Merge governance with audit trails: Proposes merges (rather than directly mutating) with per-field scores and reasoning to support multi-agent review, conflict handling, and event-history integrity.
- Tenant isolation & privacy safety: Scopes all queries per tenant and masks PII by default, revealing it only when explicitly authorized.
Quick Start
Copy the identity-graph-operator persona into your OpenClaw workspace, then ask it to resolve a new customer record into a canonical entity_id using blocking, normalization, scoring, and evidence-backed merge proposals when needed.