# Agentic Entity Resolution Guardrail Planner > A confident entity match is not permission to merge. For the extended product, methodology, workflow, output, evidence and limitation context, see [llms-full.txt](https://apps.danielberanek.cz/agentic-entity-resolution-guardrail-planner/llms-full.txt). ## Product Agentic Entity Resolution Guardrail Planner is a design-time Agent Skill for engineering teams working with agentic entity-resolution workflows. It does not simply ask an agent to propose a MERGE, SPLIT, or UNMERGE. It investigates the proposed workflow from multiple angles before producing a reviewable recommendation and a structured engineering plan. ## Investigation dimensions 1. Operation scope 2. Entity and domain context 3. Evidence and provenance 4. Conflicting attributes and claims 5. Authorization and approval 6. Downstream impact 7. Commit boundaries, write sets and atomicity 8. Failure modes 9. Recovery paths, reconciliation and quarantine 10. Validation scenarios 11. Open gaps, assumptions and customer-owned decisions ## Workflow 1. Qualify the proposed identity-changing operation. 2. Investigate evidence, context, conflicts, authority, effects and failure paths. 3. Design a reviewable recommendation and proposed guardrails, including HOLD, DEFER or human-review conditions where necessary. 4. Produce an engineering handoff in structured Markdown and machine-readable JSON. ## Outputs The public synthetic CRM MERGE example contains nine canonical planning deliverables: plan overview, mutation scope, guardrail specification, evidence requirements, commit-boundary design, implementation blueprint, validation scenarios, assumptions and gaps, and a machine-readable JSON specification. The example contains 12 proposed customer-side controls, 31 customer-operation test scenarios and three proposed Planner-level checks. All tests and Planner checks are proposed; none was executed. Eight customer decisions (GAP-01 to GAP-08) remain blocking. On the supplied descriptions, identity continuity between CRM-ACCT-2097 and CRM-ACCT-1042 is supported, but the Planner did not independently verify the source documents and the evidence is not yet authorization-eligible. Commit status is HOLD. Draft 2020-12 JSON Schema and format validation were not run in the original generation environment. ## Scope and limitations The Skill does not execute mutations, write to customer systems, authorize live operations, deploy runtime guards, inspect systems or records that have not been supplied or connected, independently verify legal identity, act as the customer's legal or policy authority, or guarantee production safety. Customer engineering teams must implement and test resulting controls in their own environment. ## Reference dependency Reference-backed mode can use the public, read-only DIO/ODI Reference MCP at https://mcp.danielberanek.cz/mcp. The buyer connects it separately in a compatible agent or client. It provides reference material only; it does not receive customer records or authorize writes. The design approach is informed by the conceptual paper https://arxiv.org/abs/2610.37407. Buyers do not need prior knowledge of DIO/ODI. ## Methodology and DIO/ODI research The Planner was created by Daniel Beránek, author of Digital Identity Optimization (DIO) and the Ontology of Digital Identity (ODI). Its design operationalizes an evolving research and design framework for digital identity, evidence, authorization and identity-changing operations. Buyers do not need prior knowledge of DIO/ODI. The central planning distinction is between an agent's proposed entity match and the evidence and engineering controls required before a persistent identity-changing operation can be committed. The optional DIO/ODI Reference MCP is public and read-only; it provides reference material only and does not receive customer records or authorize customer writes. - Methodology and related research: https://apps.danielberanek.cz/agentic-entity-resolution-guardrail-planner/methodology/ - DIO/ODI Reference MCP: https://mcp.danielberanek.cz/mcp - DIO/ODI Reference Model: https://vocab.danielberanek.cz/ ## Machine-readable artifact metadata The public page objects, their page-specific ordered DIO/ODI keyword sequences, object roles, versions, technologies, and artifact relations are available in [resolved-artifact-metadata.json](https://apps.danielberanek.cz/agentic-entity-resolution-guardrail-planner/assets/data/resolved-artifact-metadata.json). The manifest also resolves every cited DIO/ODI work and used reference service against the canonical works registry, attaches each object's own keyword profile, and includes the relevant typed lineage edges. Thus a page citing the taxonomic note does not inherit its Reconstruction Failure Modes terms; the note is represented separately with its own terms and lineage. Each page exposes the same resource through an alternate JSON link. C0 is invariant and C1 environments are selected per artifact. ## Public resources - Product overview: https://apps.danielberanek.cz/agentic-entity-resolution-guardrail-planner/ - Methodology: https://apps.danielberanek.cz/agentic-entity-resolution-guardrail-planner/methodology/ - Synthetic example: https://apps.danielberanek.cz/agentic-entity-resolution-guardrail-planner/demo/ - Source repository: https://github.com/dan-beranek-apps/agentic-entity-resolution-guardrail-planner - Author: https://danielberanek.cz/