• Skip to primary navigation
  • Skip to main content

WebMEM™

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

  • Primer
  • Memory-First
  • Protocols
    • SDT Specification
    • WebMEM SemanticMap
    • WebMEM MapPointer
    • Digest Endpoint Specification
    • ProvenanceMeta Specification
    • AI Retrieval Feedback Loop Specification
    • Semantic Feedback Interface (SFI) Specification
    • Glossary Term Protocol (GTP) Specification
    • Examples
  • RFC
  • Glossary
  • About
    • WebMEM License
    • Mission
    • Charter

WebMEM RFCs

Last Updated: 2025-07-29
Maintainer: David W. Bynon
License: CC BY-SA 4.0


About This Registry

The WebMEM™ RFC series defines the open specifications that power structured memory publishing, trust-scored retrieval, and agentic system interoperability. Each RFC represents a formal module within the protocol—ranging from fragment schema design to provenance scaffolding, glossary vocabularies, and feedback conditioning interfaces.

These specifications are modular, machine-ingestible, and open by design. They form the canonical core of the WebMEM™ Protocol and serve as public infrastructure for trustworthy AI memory, glossary federation, and context-aware retrieval systems.


Current RFCs

  1. RFC-001: Semantic Data Definition Specification
    Defines the fragment-level data_id schema for structured memory objects used in AI retrievability and dataset exposure.
  2. RFC-002: Semantic Digest Document Specification
    Establishes the structure, scope, and export requirements for full digest documents across Markdown, YAML, JSON-LD, and TTL.
  3. RFC-003: Provenance Mapping Specification
    Declares dataset and authorship lineage using W3C PROV-compatible trust blocks for memory scaffolding and citation scoring.
  4. RFC-004: Glossary Vocabulary and DefinedTermSet Specification
    Standardizes how glossary terms are defined, versioned, aligned, and published for semantic retrieval and reference resolution.
  5. RFC-005: Trust Score Computation Specification
    Outlines how memory fragments are trust-scored based on provenance depth, glossary alignment, semantic explainability, and multi-format retrievability.
  6. RFC-006: Semantic Feedback Interface Fragment Registry
    Defines the structure and retrieval model for explainers, policies, Q&A blocks, and instructional fragments used in memory conditioning workflows.

Contributing RFCs

WebMEM welcomes proposals from independent developers, publishers, researchers, and public institutions. To contribute a new RFC or suggest a revision:

  • Submit via GitHub PR: github.com/WebMEM/rfc
  • Email proposals: rfc@webmem.com
  • Optional: Archive your draft on Zenodo and include a DOI in your submission

All RFCs must include a changelog, use the RFC License (CC BY-SA 4.0), and be published in at least YAML or Markdown. JSON-LD, TTL, and PROV support are optional but encouraged.


Upcoming RFC Modules

  • RFC-007: Retrieval Confirmation Logging Interface (RCLI)
  • RFC-008: Co-Citation Scoring Schema
  • RFC-009: Trust Node Verification Protocol

The full registry is version-controlled at github.com/WebMEM/rfc.


WebMem.com is the canonical publication source for all WebMEM™ RFCs. Each module defines a formal trust-layer component for structured AI memory in agentic, interoperable systems.

Copyright © 2025 · David Bynon · Log in