What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

Agent Memory as a File Format

Connections

Supports

Serve Markdown to AI Agents with Accept Headers Serving Markdown to AI agents via Accept headers is complementary infrastructure to Memoryfields' file-based Markdown memory approach - both treat Markdown as a first-class data format for agent interaction
Managing Agent Skills with Dr. Skill Dr. Skill manages agent skill persistence; Memoryfields' file-format approach to memory provides a complementary data layer that could underpin how agent skills/experiences are stored and retrieved

Related to

6 RAG Architectures RAG architectures and Memoryfields both address the problem of external memory/retrieval for AI agents; Memoryfields argues for simpler file+vector-index approach vs. the complex multi-layer RAG architectures
WikiSkill: Agent Experience into Persistent Knowledge Both address persistent agent memory/knowledge: WikiSkill stores agent experiences as persistent knowledge, while Memoryfields proposes a file-based format for agent memory - both treat memory as structured, reusable data

Challenges

Keep AI Out of Your Obsidian Vault 'Keep AI Out of Your Obsidian Vault' argues against mixing AI with personal knowledge stores, while Memoryfields proposes embedding agent memory directly into file-based (Markdown) structures - opposing philosophies on AI/file integration

Develops into

Understanding Embeddings in Language Models Understanding embeddings is foundational to Memoryfields' 'optional vector indexes' component - the proposal builds on embedding concepts to create a hybrid file+semantic memory format

Supported by

Harness Engineering Paper Collection Agent Memory as a File Format addresses a core harness infrastructure concern—how persistent state is managed around model weights—matching the collection's focus on surrounding infrastructure evolution
Skillbay: AI Skills Marketplace Agent Memory as a File Format establishes the principle of externalizing agent capabilities into portable file-based formats — Skillbay's SKILL.md packages instantiate this exact approach for teachable coding skills
Pizza Bot - Local-First AI Agent Inbox Pizza Bot's local-first inbox for persistent agent tasks directly complements Agent Memory as a File Format - both tackle the problem of maintaining agent state across long-running interactions

Type:	link
First published:	Monday, 31 August 2026 at 19:11 UTC
Location:	https://calpaterson.com/memoryfields.html
URL:	https://mattwood.fyi/i/a3842058

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections