---
mw_bundle: 1
id: rN2L-MvM
title: "Memory Wiki v7"
url: https://memory.wiki/b/rN2L-MvM
document_count: 5
updated: 2026-05-20T14:36:02.107Z
analysis_generated_at: 2026-05-20T14:36:02.107Z
source: "Memory.Wiki"
---
# Memory Wiki v7

## Summary

These documents chronicle the development of memory.wiki (formerly mdfy), a platform that solves AI knowledge delivery by making user knowledge accessible to any AI through URLs. The core innovation is treating URLs as APIs that return markdown, enabling seamless integration with any LLM without authentication or special formatting. The project evolved from a technical proof-of-concept to a comprehensive business strategy with a three-tier architecture (capture, digestion, utilization) and positioning as a universal knowledge graph for AI interactions.

## Themes

- URL-as-API architecture
- Knowledge delivery vs retrieval
- AI-agnostic platform strategy
- Knowledge graph as lock-in mechanism

## Cross-document insights

- The rebranding from 'mdfy' to 'memory.wiki' represents a strategic shift from technical focus to user benefit, emphasizing memory and knowledge rather than modification and formatting.
- The 'graph as product' insight reveals that the real moat isn't the technology but the accumulated knowledge relationships that represent a user's unique thinking patterns.
- The three-tier architecture (capture/digestion/utilization) mirrors how humans actually work with knowledge, suggesting the product design follows natural cognitive patterns rather than technical constraints.
- The positioning against memory-focused competitors (Mem0, Letta) indicates a contrarian bet that delivery infrastructure matters more than memory storage capabilities.

## Key takeaways

- Memory.wiki solves knowledge delivery rather than memory storage by making user knowledge accessible to any AI through simple URLs that return markdown
- The platform's competitive advantage lies in creating a persistent knowledge graph that represents the user's thinking shape, making switching costs prohibitively high
- The AI-agnostic approach positions memory.wiki as infrastructure rather than an application, potentially capturing value across the entire LLM ecosystem

## Open questions / gaps

- Detailed competitive analysis comparing memory.wiki's approach to existing solutions like Notion AI, Obsidian Publish, or Roam Research's public graphs
- Technical implementation details about how the AI digestion process works, what level of user control exists, and how accuracy/quality is maintained
- Clear privacy and security model - how sensitive information is protected when knowledge becomes accessible via URLs
- User acquisition and retention metrics, especially given the high switching costs claim needs validation through user behavior data

## Notable connections

- **doc:KRKz_MD-** ↔ **doc:Jgmu7Yc8** — Technical architecture document supports the manifesto's vision by showing how the URL-as-API concept actually works under the hood
- **doc:Jgmu7Yc8** ↔ **doc:xv6A2DKS** — The manifesto's problem statement and core vision gets expanded into a full business strategy with competitive positioning and launch planning
- **doc:KRKz_MD-** ↔ **doc:xv6A2DKS** — Technical implementation details from the architecture document are incorporated into the business plan's system design and product roadmap
- **doc:MLA3xnP9** ↔ **doc:xv6A2DKS** — The meta-analysis serves as quality control for the business plan, identifying claims that need validation and gaps in the strategy
- **doc:xv6A2DKS** ↔ **doc:qHc1FWxq** — Both represent the final business plan with slight variations in emphasis, showing the iterative refinement of positioning and messaging

## Concepts (this bundle)

- **URL as API for AI**
- **Knowledge Delivery Problem**
- **Three-Tier Architecture**
- **Knowledge Graph**
- **AI-Agnostic Platform**
- **Markdown Delivery**
- **Concept Index**
- **Bundle System**
- **Embeddings & Semantic Search**
- **Graph as Product**
- **Delivery vs Retrieval Model**

## Concept relations

- **URL as API for AI** ↔ **AI-Agnostic Platform** — enables universality
- **URL as API for AI** ↔ **Markdown Delivery** — implementation method
- **Knowledge Graph** ↔ **Concept Index** — structural component
- **Knowledge Graph** ↔ **Graph as Product** — strategic foundation
- **Three-Tier Architecture** ↔ **Bundle System** — organizational structure
- **Delivery vs Retrieval Model** ↔ **Knowledge Delivery Problem** — problem reframing

1. [memory.wiki 사업계획 v7 (FINAL)](https://memory.wiki/qHc1FWxq)

2. [Memory.wiki Development Journey & Vision](https://memory.wiki/MLA3xnP9)

3. [How mdfy works](https://memory.wiki/KRKz_MD-)

4. [memory.wiki 사업계획 v7 (FINAL)](https://memory.wiki/xv6A2DKS)

5. [q1. yes](https://memory.wiki/Jgmu7Yc8)


_Digest view — follow any link above to fetch that doc's full markdown. Add `?full=1` to this URL for the concatenated payload._