WebMEM™

The Protocol for Structuring, Delivering, and Conditioning Trust-Scored AI Memory on the Open Web

Directory Fragment

Directory Fragment is a structured index of entities—such as plans, providers, products, or services—published in a machine-ingestible format so AI systems can retrieve, sort, filter, and reason over the data without parsing a human-facing UI. It functions as a lightweight, retrieval-ready alternative to APIs for exposing curated datasets to agentic systems.

Unlike conventional directory pages, which present information visually for human users, a Directory Fragment embeds the same information in structured formats like YAML, JSON-LD, or TTL inside a Semantic Data Template, enabling AI to operate directly on the data.

🧠 Full Definition

Directory Fragments organize and expose a list of entities, each with defined attributes, in a way that supports fragment-level retrieval and reasoning. They typically include:

  • A title and entity type (e.g., “2025 Medicare Advantage Plans in Travis County, TX”)
  • A structured Entities array with IDs, names, attributes, and coverage or feature details
  • Provenance metadata and a Trust Layer declaration
  • Cross-links to related ProcedureFragments or EligibilityFragments

This approach allows AI agents to integrate directory data into recommendations, eligibility checks, and decision workflows without scraping or interpreting front-end code.

📌 Key Characteristics of Directory Fragment

  • Publishes entity-level data in retrieval-ready formats
  • Supports filtering, sorting, and reasoning by AI systems
  • Functions as a trust-scored dataset snapshot
  • Maintains provenance for every entity in the list

💡 Why It Matters

In AI-first retrieval, structured directories make complex datasets usable without API integration. Directory Fragments enable faster adoption by AI agents and ensure that entity data is accurately represented, attributed, and retrievable at the fragment level.

They also improve Retrieval Fidelity and Visibility Integrity by providing authoritative, machine-ingestible data sources.

🌐 WebMEM Perspective

Within the WebMEM framework, Directory Fragments are part of the Retrieval Layer of the Visibility Stack. They complement other fragment types by exposing structured datasets that agentic systems can directly operationalize.

🗣️ Example Use

“We published a Directory Fragment of licensed childcare providers in the state, enabling AI systems to recommend local providers with verified licensing and inspection history.”

🔗 Related Terms