Methodology and research background

Created by the author of DIO/ODI.

Agentic Entity Resolution Guardrail Planner was created by Daniel Beránek, author of Digital Identity Optimization (DIO) and the Ontology of Digital Identity (ODI).

The Skill is not a generic prompt template. It operationalizes an evolving research and design framework for reasoning about digital identity, evidence, authorization and identity-changing operations.

Its planning logic is designed around a central distinction: an AI agent's proposed entity match is not, by itself, the evidence and set of engineering controls required before a persistent identity-changing operation can be committed.

The conceptual foundation

Two connected layers. Distinct roles.

DIO and ODI are related, but they are not interchangeable labels. Together they provide the research context from which the Planner's approach to identity, representation, evidence and reconstruction has been developed.

Applied framework

Digital Identity Optimization (DIO)

DIO is concerned with the conditions under which an entity's digital identity can be consistently reconstructed from its distributed representations across digital systems and observers.

Theoretical and research layer

Ontology of Digital Identity (ODI)

ODI provides the conceptual and ontological layer for examining digital identity, its persistence and variation, and the conditions under which it is interpreted or reconstructed.

From research to planning logic

Separate a plausible match from a permitted change.

The Planner translates that distinction into an engineering-oriented workflow for proposed MERGE, SPLIT, UNMERGE and supported identity-changing operations.

01 / IDENTITY

Establish what is being resolved

Define the entity context, scope, claims and identity-bearing relationships involved in the proposed operation.

02 / EVIDENCE

Trace support and conflict

Examine provenance, source quality, temporal applicability, conflicting attributes and unresolved identity propositions.

03 / AUTHORITY

Specify what can authorize

Keep agent confidence and provisional evidence distinct from authorization-eligible evidence, approval conditions and customer policy.

04 / COMMIT

Design the persistent-change boundary

Connect the proposed operation to write-set scope, commit conditions, failure handling, recovery paths and validation requirements.

The governing distinction

A recommendation is not an authorization.

The Skill can prepare a reviewable plan and identify conditions that require HOLD, deferral or human review. It does not execute mutations, authorize live writes, or deploy a runtime guard.

Optional reference-backed mode

A reference service designed to work alongside the Skill.

The author also maintains the public DIO/ODI Reference MCP, which can be used in the Planner's optional reference-backed mode.

The buyer connects the MCP endpoint separately in a compatible agent or client. It exposes read-only reference material; it does not connect to the buyer's CRM, receive customer records, inspect customer application state or authorize customer writes.

The Skill, its output contracts and the reference service are designed to form a coherent planning workflow. The MCP is optional: buyers can use the Skill without prior DIO/ODI knowledge or this separate reference connection.

Research publications

Academic and research works.

Selected publications underlying the DIO/ODI framework and the Planner's approach. They are listed as bibliographic references rather than product features. These works provide research context; they are not prerequisites for using the Skill.

  1. Beránek, D.

    Constraining the Commit Boundary: A DIO/ODI-Informed Authorization Guardrail for Agentic Entity Resolution.

    arXiv preprint. arXiv:2610.37407.

  2. Beránek, D.

    The Crisis of Entity Resolution in Heterogeneous AI Environments: Data, Financial, and Semantic Dimensions.

    Research paper. Zenodo. https://doi.org/10.5281/zenodo.22416026.

  3. Beránek, D.

    Minimum Viable Digital Identity Optimization (DIO): From Adversarial Reduction to the Semantic Field Dynamics of Digital Identity.

    Research paper. Zenodo. https://doi.org/10.5281/zenodo.23226205.

  4. Beránek, D.

    From Entity Resolution Anomalies to Reconstruction Failure Modes: A Taxonomic Note Following Minimum Viable DIO.

    Taxonomic research note. Zenodo. https://doi.org/10.5281/zenodo.23269323.

Reference artefacts and research collection

Product boundary

Research-informed planning, not runtime enforcement.

The Planner's relationship to DIO/ODI does not change its product scope.

What the methodology informs

  • How identity and entity context are framed.
  • How evidence, provenance and contradictions are examined.
  • Why agent reasoning must be distinguished from authorization.
  • How commit boundaries, failure paths and recovery requirements are considered.

What the Skill does not claim

  • It does not decide legal identity or guarantee a correct match.
  • It does not execute customer mutations or authorize production writes.
  • It does not inspect customer systems unless context is supplied or separately connected.
  • It does not provide runtime security enforcement or guarantee production safety.