--- title: "mdfy.cc Manifesto" url: https://memory.wiki/3Ug2sp5E updated: 2026-04-26T22:55:20.970Z hub: https://memory.wiki/hub/raymindai concept_count: 12 source: "desktop" --- # mdfy.cc Manifesto > 지금까지 모든 분석을 거쳐 확정한 방향. 모든 후속 작업의 anchor 문서. ⌘Enter to exit block * * * ## 1\. Manifesto ### 한 줄 정의 > **Own your AI memory. Use it anywhere.** ⌘Enter to exit block ### 핵심 가치 mdfy는 AI 시대의 메모리에 대한 **철학적 입장**을 가진 product다. **우리가 믿는 것**: 1. **Memory is yours, not extracted.** ⌘Enter to exit block - Mem0/Letta처럼 AI가 chat에서 자동 추출하는 memory는 좋은 도구지만, 다른 종류의 memory다. - mdfy는 사용자가 **의도적으로 선택**한 것을 저장한다. ⌘Enter to exit block 1. **Memory should be portable.** ⌘Enter to exit block - Markdown은 AI 시대의 lingua franca. - URL은 가장 단순한 interface. - SDK lock-in 없음. 어떤 AI와도 호환. ⌘Enter to exit block 1. **Memory deserves intention.** ⌘Enter to exit block - 무엇이 들어가는지가 무엇이 되는지를 결정한다. - 이건 사람의 memory에도, AI의 memory에도 적용된다. ⌘Enter to exit block 1. **Memory should be readable.** ⌘Enter to exit block - 사람이 브라우저에서 그대로 읽을 수 있어야 한다. - AI가 URL로 fetch 가능해야 한다. - 둘 다 같은 markdown 문서를 본다. ⌘Enter to exit block * * * ## 2\. Hero (확정) **H1**: Own your AI memory. Use it anywhere. **Sub**: Capture answers from ChatGPT, Claude, Cursor. Edit them in markdown. Paste them back into any AI as context. Your knowledge — owned, edited, portable. **CTA primary**: Start your memory → **CTA secondary**: Install Chrome extension * * * ## 3\. Strategic Position ### 시장에서의 자리 ``` AI extracts You author (자동 학습) (의도적 작성) ↓ ↓ LLM-native Mem0, Letta [mdfy의 자리] ⭐ OpenAI Memory Google Memory Bank Human-friendly Notion AI Obsidian, GitHub Gist, Coda HackMD ``` ⌘Enter to exit block 비어있는 자리: **LLM-native + 사용자가 직접 author + open standard**. ### 차별화 narrative > “Mem0 is excellent at what it does — it extracts memory from your conversations. > > mdfy is different — you author what your AI remembers.” ⌘Enter to exit block **정직한 포지셔닝**: - Mem0/Letta: AI extract → conversation memory에 적합 - mdfy: User author → curated knowledge에 적합 - 경쟁자 아닌 보완재. 함께 사용 가능. ⌘Enter to exit block ### 진짜 경쟁자 (정직하게) - **정면**: GitHub Gist, HackMD (markdown URL publishing) - **인접**: Notion, Obsidian Publish (PKM) - **보완재**: Mem0, Letta (다른 layer의 memory) ⌘Enter to exit block * * * ## 4\. The Five Pillars of mdfy 이전 분석에서 확인된 mdfy의 5가지 진짜 강점: 1. **Author-first** — 사람이 직접 작성한다 2. **Document primitive** — 분해 안 하고 사람이 읽는 단위 그대로 3. **URL composability** — 어디든 붙여넣기 가능 4. **Multi-surface** — CLI, MCP, Web, VS Code, Mac, Chrome 모두 build됨 5. **Open standard** — markdown 자체가 표준 ⌘Enter to exit block 이 5가지가 marketing의 모든 메시지를 정의한다. * * * ## 5\. Product Wedge: AI Answer Collection-first ### Wedge 정의 **진입점**: AI 답변을 수집해서 영구 URL로 저장하는 것. **Why this wedge**: - 매일 일어나는 pain (모든 ChatGPT/Claude 사용자) - 즉시 understandable (1초 안에 GETS IT) - Viral 강함 (Chrome extension은 word-of-mouth 빠름) - mdfy의 multi-surface 자산이 정확히 fit - Memory infra 비전으로 자연 진화 ⌘Enter to exit block ### 자연 funnel ``` Layer 1: Collection (entry, free, viral) "AI 답변 어디서나 한 번에 저장" ↓ Layer 2: Curation (Pro 전환 시작) "수집한 답변을 정리, 편집, 공유" ↓ Layer 3: AI integration (Build 전환) "당신의 markdown을 AI가 읽고 씀" ``` ⌘Enter to exit block * * * ## 6\. Phase 구조 | Phase | 기간 | Primary 메시지 | Hidden 자산 | | --- | --- | --- | --- | | Phase 1 | Month 1-3 | “Save AI answers to permanent URLs” | Memory infra (silent) | | Phase 2 | Month 4-6 | \+ “Curate your AI knowledge” | AI integration 활성화 | | Phase 3 | Month 7-9 | \+ “Use it as context anywhere” | Team docs (자연 발생) | | Phase 4 | Month 10-12 | “The memory layer for AI-native work” | Enterprise + standard | 핵심 전환 trigger: - Phase 1 → 2: Chrome ext 설치 1만+ / DAU 500+ - Phase 2 → 3: $5K MRR + Build tier paid 100명+ - Phase 3 → 4: $20K MRR + 팀 워크스페이스 30개+ ⌘Enter to exit block * * * ## 7\. Pricing 구조 ``` Free - Unlimited collection from any AI - 100 saved answers - Basic edit/share - Public URLs only - Community MCP server Pro $9/mo (개인) - Unlimited saved answers - Private URLs - Folders, search, tags - Custom domain - Sync across devices - Version history Build $19/mo (AI 빌더) - API access (read/write) - MCP server (full) - Paste back to AI as context (자동화) - Webhook integrations - Embeddings & semantic query - Agent definitions Team $19/seat/mo (팀) - Shared collections - Team workspace - Permissions, audit log - SSO (Google) - Slack integration Scale $499+/mo (heavy infra) - Custom rate limits - SLA, priority support - Multi-region - Dedicated MCP Enterprise (협의) - Self-host option - SAML SSO - Custom audit/compliance - Dedicated support ``` ⌘Enter to exit block 핵심 monetization driver: - **Phase 1-2**: Pro $9 (consumer, viral) - **Phase 2-3**: Build $19 (power user, AI integration) - **Phase 3-4**: Team + Enterprise ⌘Enter to exit block * * * ## 8\. Three Pillars (landing page용) Hero 아래 들어갈 3개 섹션: ### 1\. Capture **From any AI, anywhere.** - Chrome extension (ChatGPT, Claude, Gemini) - VS Code extension - Mac app (clipboard watch) - CLI (`cat | mdfy`) - Direct paste, file upload ⌘Enter to exit block ### 2\. Edit **In beautiful markdown WYSIWYG.** - Your voice, your structure - Version history, diff - AI-assisted polish (optional) - Privacy: public, unlisted, or private ⌘Enter to exit block ### 3\. Use **As context in any AI conversation.** - Permanent URL: paste anywhere - MCP server: AI agents fetch automatically - API: programmatic access - Read by humans in browser ⌘Enter to exit block * * * ## 9\. 12개월 KPI 목표 ### Month 3 (Phase 1 종료) - Chrome extension 설치 1만+ - DAU 500+ - 5만 saved answers - $1K MRR (early Pro adopters) ⌘Enter to exit block ### Month 6 (Phase 2 종료) - 사용자 5만+ - DAU 2,500+ - 30만 saved answers - $10K MRR - Build tier 활성화 시작 ⌘Enter to exit block ### Month 9 (Phase 3 종료) - 사용자 15만+ - $30K MRR - AI agent integrations 20+ - Team workspace 30+ ⌘Enter to exit block ### Month 12 (Phase 4 진입) - $50-80K MRR - 첫 enterprise pilot 1-2건 - 표준 인지도 (industry mention) - Strategic conversation 시작 ⌘Enter to exit block * * * ## 10\. 차별화 메시지 (정리) ### 강한 메시지 (사용) ✅ **“Author your own AI memory”** — 핵심 가치 ✅ **“Owned, edited, portable”** — 3단어 manifesto ✅ **“From AI chat to permanent URL”** — workflow 명확 ✅ **“Your knowledge layer”** — ownership ✅ **“Markdown URLs that humans write and AI reads”** — composability ### 약한 메시지 (피함) ❌ “Black-box memory” (부정확 — Mem0도 OSS) ❌ “Vendor lock-in” (부정확 — multi-provider 호환) ❌ “Don’t let AI extract you” (부정형, paranoid 톤) ❌ “Write yourself” (manual labor 인상) ❌ “Markdown Hub” (commodity, 차별화 0) * * * ## 11\. Operational Plan ### 시간 배분 - **Month 1-3**: 80%+ (foundation + Chrome ext launch) - **Month 4-6**: 70% (curation tier + viral 가속화) - **Month 7-9**: 60% (AI integration + 팀 자연 발생) - **Month 10-12**: 50%+ (Scale + Enterprise inquiry) ⌘Enter to exit block ### 자동화 우선순위 1. Self-serve 결제 (Stripe Checkout, 모든 tier) 2. AI customer support (Claude로 1차 응대) 3. Onboarding 자동화 (email sequence, in-app tutorial) 4. Dev support (좋은 docs, GitHub issues, Discord) 5. Content marketing (AI draft + 직접 polish) ⌘Enter to exit block ### 외주 검토 - Month 6: 영문 카피 polish ($300-500) - Month 9: Part-time content writer ($1-2K/월) - Month 12: 첫 풀타임 hire 검토 ⌘Enter to exit block * * * ## 12\. Risk Register ### Risk 1: 거인들 진입 (확률 60%, impact 매우 큼) - Anthropic의 official memory layer - OpenAI Memory 확장 - Google Memory Bank ⌘Enter to exit block **Mitigation**: - Open spec 빠르게 발표 - Multi-LLM agnostic 강조 - Community 기반 OSS positioning - 빠른 launch (6개월 안에 인지도) ⌘Enter to exit block ### Risk 2: Build에 빠져 launch 지연 (확률 70%, impact 매우 큼) - 가장 큰 위험 — 1년 build해놓고 launch 안 함 ⌘Enter to exit block **Mitigation**: - Month 3 beta launch hard commit - Month 5 public launch hard commit - “Perfect는 launch 후” 원칙 ⌘Enter to exit block ### Risk 3: 메시지 dilution (확률 60%, impact 중간) - “Hybrid” 이름으로 메시지 흩어지는 함정 ⌘Enter to exit block **Mitigation**: - Manifesto 절대 안 바뀜 - Phase 1은 single message (Collection) - 매 분기 메시지 audit ⌘Enter to exit block ### Risk 4: Memory/Harness 시장이 너무 일찍 (확률 30%) - 2026년에 production agent 부족 → 매출 정체 ⌘Enter to exit block **Mitigation**: - Phase 1 Collection wedge가 시장 무관하게 작동 - Memory infra는 자연 확장 - 12개월 후 traction 보고 pivot 가능 ⌘Enter to exit block ### Risk 5: 솔로 운영 burnout (확률 50%, impact 매우 큼) **Mitigation**: - Month 6부터 자동화 강화 - Month 9부터 part-time 외주 - 매주 1일 OFF 강제 ⌘Enter to exit block * * * ## 13\. 즉시 다음 Task (우선순위) ### Week 1 (이번 주) **A. Three pillars 본문 작성** (1-2시간) ← 다음 task - Capture / Edit / Use 각 섹션 헤드라인 + 본문 ⌘Enter to exit block **B. Demo 영상 3편 시나리오** (1시간) - Capture: Chrome ext에서 ChatGPT 답변 저장 - Edit: WYSIWYG에서 polish - Use: Cursor/Claude에 paste ⌘Enter to exit block **C. About 페이지 재구성** (2-3시간) - Manifesto 톤 적용 - 4-belief 본문 - Product description ⌘Enter to exit block ### Week 2 **D. Manifesto post 초안** (3-4시간) - HN/Substack용 1500-2500 단어 - “Own your AI memory” 철학 풀어내기 - Hyunsang님 personal voice ⌘Enter to exit block **E. Pricing 페이지** (1-2시간) - 6 tier 카피 finalize ⌘Enter to exit block **F. API/MCP 문서 정비** - Build tier 사용자 onboarding ⌘Enter to exit block ### Week 3-4: Beta 준비 **G. Chrome ext UX 점검** (viral 메커니즘) **H. Beta tester 50명 모집 plan** **I. Launch assets** (Show HN 글, PH 페이지, demo gif 등) ### Month 2-3: Beta + iteration **J. 50명 alpha tester 모집 + onboard** **K. 매주 피드백 → 빠른 iteration** **L. 5-10 case study 정리** ### Month 4-5: Public Launch **M. HN Show HN, Product Hunt, Twitter** **N. Manifesto post 공식 발표** **O. AI dev community marketing** * * * ## 14\. 핵심 결정사항 (확정) | 결정 | 답 | | --- | --- | | Manifesto | “Own your AI memory. Use it anywhere.” | | Wedge | AI Answer Collection-first | | 진입 surface | Chrome extension | | 첫 launch 채널 | HN Show HN | | Pricing entry | Free → Pro $9 | | 차별화 frame | Authorship (User authors vs AI extracts) | | 경쟁자 인식 | GitHub Gist 정면, Mem0 보완재 | | Open source | mdcore OSS, mdfy.cc SaaS | | Phase 구조 | Collection → Curation → AI integration → Team | | 시간 commit | Month 1-3 80%+ | * * * ## 15\. 한 줄로 > mdfy는 사용자가 직접 author하는 AI memory의 layer다. > > 어디서든 capture, markdown으로 edit, 어디서든 use. > > Owned, edited, portable. ⌘Enter to exit block 이 한 단락이 흐려지면 사업이 흐려진다. * * * _Last updated: 2026-04-27_ > 이렇게 된듯 아닌듯?1 --- ## Summary mdfy is a product that lets users author and own their own AI memory in markdown format, capturing answers from any AI tool, editing them, and using them as portable context anywhere, positioning itself as user-authored curation rather than AI-extracted memory like competing services. ## Themes - User authorship over AI extraction - Portable markdown-based memory - Multi-surface AI integration ## Key takeaways - mdfy's core value proposition is that memory ownership belongs to the user who intentionally authors it, not to AI systems that automatically extract from conversations. - The product enters through AI answer collection via Chrome extension but naturally evolves into a memory infrastructure layer supporting AI integration and team workspaces. - Five pillars define mdfy's competitive advantages: author-first creation, document-level primitives, URL composability, multi-surface availability, and open markdown standards. - The 12-month roadmap has four distinct phases, each unlocking a higher monetization tier: Phase 1 is free viral collection, Phase 2 introduces Pro curation, Phase 3 enables Build-tier AI integration, Phase 4 targets Enterprise and standardi - Three real competitors are identified honestly: GitHub Gist and HackMD as direct markdown URL alternatives, Notion/Obsidian as adjacent PKM tools, and Mem0/Letta as complementary memory layers. ## Insights - The positioning deliberately avoids direct competition with Mem0/Letta by occupying a different layer (user-curated vs AI-extracted), treating them as complementary rather than rival. - Launch delay is identified as the highest-impact risk (70% probability), suggesting the founder recognizes overbuilding as a greater threat than market competition. - The pricing funnel mirrors the product maturation: free collection drives adoption, Pro ($9) captures individual users, Build ($19) unlocks AI integration revenue, and Team/Enterprise captures organizational value. ## Open questions / gaps - How will mdfy prevent the documented Risk 1 scenario where major LLM providers (Anthropic, OpenAI, Google) launch competing memory layers with native platform advantages? - What specific metrics or user behaviors will trigger the Phase transition decision points (e.g., how is DAU 500 measured, and what if it reaches 450?) ## Concepts in this document - **Chrome extension** _(entity)_ Surface within memory.wiki ecosystem. - **Knowledge Management** _(tag)_ Overarching domain of personal and organizational information systems - **mdfy** _(entity)_ A memory infrastructure layer that provides a URL-based knowledge hub for AI tools. - **Claude** _(entity)_ Specified AI tool for prototyping and validation before moving to high-fidelity design. - **ChatGPT** _(entity)_ One of the AI platforms currently suffering from isolated memory silos. - **URL Architecture** _(concept)_ Shared URL structure between web and iOS to maintain canonical document and profile access. - **Model Context Protocol (MCP)** _(entity)_ Technical protocol enabling AI agents to interact with mdfy documents for memory management. - **Cross-AI Compatibility** _(concept)_ The ability for Memory.Wiki URLs to work across any AI tool without vendor lock-in - **Markdown** _(tag)_ Lightweight markup format used as the universal content format across Memory.Wiki. - **memory portability** _(concept)_ The core principle that memory should follow the user rather than being locked into specific AI models. - **Obsidian** _(entity)_ Competitor noted as a note-taking tool in relation to memory concepts. - **Mem0** _(entity)_ Competitive reference mentioned in the redefinition document. ## Concept relations (within this doc's concepts) - **Markdown** enables cross-platform use **memory portability** - **mdfy** provides platform **Knowledge Management** - **mdfy** distributed via **Chrome extension** - **mdfy** integrates with **ChatGPT** - **mdfy** integrates with **Claude** - **Chrome extension** captures content for **mdfy** - **Model Context Protocol (MCP)** bridges ai ecosystems **mdfy** _Hub canonical:_ https://memory.wiki/hub/raymindai _Concept digest:_ https://memory.wiki/raw/hub/raymindai?digest=1&compact=1