--- title: "mdfy.cc 사업화 방향성 v4 (FINAL)" url: https://memory.wiki/-2UojXG0 updated: 2026-04-26T22:53:45.227Z hub: https://memory.wiki/hub/raymindai concept_count: 12 source: "desktop" --- # mdfy.cc 사업화 방향성 v4 (FINAL) > Broad markdown tool로 진입, Memory layer로 진화. 8주 launch commit. Standard-setting movement. --- ## 0. Founder Context - **Build 기간**: 부업 1개월 (압도적으로 빠른 velocity) - **현재 상태**: 풀타임 mdfy commit, 사용자 0명, 마케팅 0 - **자원 제약**: 없음 (자본/시간 압박 없음) - **Ambition**: Big — standard-setting movement --- ## 1. Core Insight (v4의 새 framing) mdfy는 두 layer의 product다: ### Layer 1: 모든 사용자를 위한 markdown 도구 - AI 답변 capture, WYSIWYG edit, permanent URL - Memory 기능 없이도 valuable - HackMD, GitHub Gist 대안 - 즉시 utility ### Layer 2: Power user를 위한 memory layer - Bundle, semantic search, MCP, API - Curated memory authored by user - AI agent integration - Long-term moat + differentiation **핵심**: 두 layer 모두 valuable. Memory는 강제 아닌 자연 진화. --- ## 2. Manifesto (별도 페이지로 deep) ### 한 줄 정의 > **Own your markdown. Use it anywhere.** ### Manifesto Belief mdfy.cc/manifesto 페이지에 이 철학을 깊이 있게: **우리가 믿는 것**: 1. **Markdown is the right primitive for AI-era knowledge.** - Plain text가 영원하다. - LLM이 native로 쓰고 읽는 format. - 어떤 도구에든 호환. 2. **URLs are the right interface.** - 가장 단순한 share 방법. - AI가 fetch 가능. - SDK lock-in 없음. 3. **Memory is yours, not extracted.** - Mem0/Letta처럼 AI가 추출하는 memory는 좋은 도구지만 다른 종류다. - mdfy는 사용자가 의도적으로 author한 것을 저장한다. 4. **Memory should be deployable.** - 저장만 하지 않고 어디서든 쓸 수 있어야 한다. - Bundle 단위로 묶고, 어떤 AI에든 컨텍스트로 deploy. 5. **Memory deserves intention.** - Bundle은 의도의 visible artifact. - Versioning은 의도의 시간적 표현. --- ## 3. Hero (확정) **H1**: Own your markdown. Use it anywhere. **Sub**: Capture AI answers. Edit in WYSIWYG. Share, bundle, deploy as context. Owned, edited, portable. **CTA primary**: Start writing → **CTA secondary**: Install Chrome extension --- ## 4. Three Pillars (Universal) ### 1. Capture **From any AI, anywhere.** - Chrome extension (ChatGPT, Claude, Gemini) - VS Code extension - Mac app, CLI, paste - Auto-source detection ### 2. Edit **In beautiful markdown WYSIWYG.** - Your voice, your structure - Tags, folders, search - AI tools (polish, summary, translate) - Version history ### 3. Use **Read, share, deploy.** - Permanent URL — works anywhere - Public, unlisted, or private - **Bundle multiple URLs as context** (Pro+) - **Read by AI via MCP/API** (Build+) → “Use”가 통합 verb. Bundle은 advanced feature로 자연 등장. --- ## 5. The Bundle — 핵심 차별화 (but backseat in Phase 1) ### Bundle의 4가지 속성 1. **Composable** — 여러 mdfy URL을 묶음, 묶음 자체도 URL 2. **Versioned** — Snapshot 가능, evolution 추적 3. **Deployable** — AI에게 한 번에 전달 4. **Curated** — 의도적 선택, annotation 포함 ### Bundle Use Cases 1. **AI conversation 시작용** — “About Me Bundle” 2. **프로젝트별 컨텍스트** — Cursor/Claude Code 즉시 deploy 3. **Agent definition** — Agent의 brain이 한 URL 4. **Knowledge sharing** — 큐레이션된 지식 공개 단위 5. **Self-context (메타)** — 시간 차원 self-archive ### Bundle의 Phase별 역할 **Phase 1 (Launch)**: - Pro tier에 포함 (5 bundles) - “Multi-select → Bundle” 자연 발견 - Hero에서 강조 안 함 - 사용자가 organic하게 발견 **Phase 2 (Launch 후 1-3개월)**: - Bundle을 second wave hero로 격상 - “Bundle your knowledge” 캠페인 - Build tier ($19) 본격 push **Phase 3 (3-6개월)**: - Public bundle library - Bundle marketplace 가능성 - Standard spec 발표 --- ## 6. Strategic Position ### 시장 자리 (재정의) ```text AI extracts You author + bundle (자동 학습) (의도적 + 묶음 deploy) ↓ ↓ LLM-native Mem0, Letta [mdfy의 Layer 2] ⭐ OpenAI Memory Google Memory Bank Markdown (없음) [mdfy의 Layer 1] ⭐ publishing HackMD, GitHub Gist Notion (closed) ``` **mdfy = 두 시장 동시 진입**: - Layer 1: Markdown publishing 시장 (HackMD, Gist 대안) - Layer 2: AI memory layer 시장 (Mem0와 다른 segment) ### 차별화 narrative (두 layer) **Layer 1 (broad)**: > “The fastest way to publish markdown for the AI era. Capture from anywhere, edit beautifully, share with a permanent URL.” **Layer 2 (deep)**: > “And when you’re ready: bundle your knowledge into deployable AI context. Own what your AI knows.” ### 경쟁자 인식 (정직하게) **Layer 1 경쟁자**: - 정면: HackMD, GitHub Gist, StackEdit - 인접: Notion (closed), Obsidian Publish (PKM 중심) - mdfy 우위: AI capture native + multi-surface **Layer 2 경쟁자**: - 정면: 없음 (Bundle 개념 unique) - 인접: Mem0/Letta (보완재, 다른 layer) - mdfy 우위: User-authored + deployable --- ## 7. Wedge: Universal Markdown Tool ### 진입점 **가장 넓은 entry**: “AI 시대의 markdown 도구” - 모든 AI heavy user 타겟 - Memory에 관심 있는 사람만 아님 - Markdown publishing이 그 자체로 valuable ### 자연 funnel ```text Layer 1: Capture (Free, viral) "AI 답변 어디서나 한 번에 저장" ↓ (모든 사용자) Layer 2: Edit/Share (Free → Pro $9) "WYSIWYG, custom domain, 검색" ↓ (regular user) Layer 3: Bundle (Pro → Build $19) ⭐ "Bundle을 발견하면서 power user 진화" ↓ (power user) Layer 4: AI Integration (Build expansion) "MCP/API로 자동화" ``` 각 layer는 독립적으로 valuable. 다음 layer로 강제 안 함. --- ## 8. Pricing (조정) ```text Free - Unlimited markdown publishing - Permanent URLs - WYSIWYG editor - Public posts - Chrome extension capture - Community MCP server (read-only) Pro $9/mo (개인) ⭐ 핵심 conversion - Private posts - Custom domain - Tags, folders, basic search - AI tools (polish, summary, translate) - 5 bundles ⭐ (memory entry) - Version history (basic) Build $19/mo (Power user) ⭐ Memory 본격 - Unlimited bundles - API access (read/write) - MCP server (full, write enabled) - Bundle versioning + snapshots - Bundle analytics - Webhook integrations - Semantic search - Format adapters Team $19/seat/mo - Shared workspace - Shared bundles - Permissions - SSO (Google) - Slack integration Scale $499+/mo - Public bundle marketplace access - Custom rate limits - SLA, priority support - Dedicated MCP Enterprise (협의) - Self-host option - SAML SSO - Custom audit/compliance ``` ### Pricing strategy 의도 **Free → Pro $9**: - Free에서 publishing 충분 가능 - Pro 진입 이유: 검색, custom domain, **5 bundles** (memory 맛보기) **Pro → Build $19**: - Bundle을 power user 정도 사용하면 5개로 부족 - Unlimited bundle + API/MCP가 진짜 차별화 - Memory layer 완전 활성화 **Notion/Linear 모델**: - 낮은 entry → mass adoption - Power user 자연 전환 - Year 2-3에 enterprise + standard으로 큰 수익 --- ## 9. The Five Pillars of mdfy (재확인) 1. **Author-first** — 사람이 직접 작성 2. **Document primitive** — 사람이 읽는 단위 그대로 3. **URL composability** — 어디든 붙여넣기 4. **Multi-surface** — CLI, MCP, Web, VS Code, Mac, Chrome 5. **Bundle deployment** — 묶어서 deploy (Phase 2 hero) --- ## 10. Launch Scope (확정) ### Tier 0: Foundation (이미 있음) - ✅ mdcore 엔진 (Rust) - ✅ mdfy.cc 웹 (publish, edit, WYSIWYG) - ✅ Permanent URL - ✅ Chrome extension (capture) - ✅ MCP server (basic read) ### Tier 1: Universal Markdown Tool **1. Tagging system** - Manual tag 추가 - Tag별 view **2. Folders/Categories** - 단순 hierarchy - 답변 이동 **3. Basic Search** - Keyword + tag 필터 **4. Privacy controls** - Public / Unlisted / Private - Pro tier에서 private 활성화 ### Tier 2: Bundle MVP **1. Bundle 생성** - Multi-select 답변 → “Create Bundle” - Bundle 이름 + description **2. Bundle 페이지** - Inline preview of referenced docs - “Copy as context” 버튼 - Source AI별 표시 **3. Bundle 편집** - 답변 add/remove - Drag-drop reorder - Annotation 추가 **4. Bundle deploy** - Copy as context (clipboard) - URL share ### Tier 2.5: Memory Excellence (Big launch) **1. Semantic Search ⭐** - OpenAI text-embedding-3-small - Hybrid 자동 (keyword + semantic) - pgvector 저장 **2. Bundle Versioning ⭐** - 명시적 snapshot 버튼 - 새 URL (mdfy.cc/bundle/abc/v1) - Live와 versions 분리 --- ## 11. Launch Roadmap (8주) ### Week 1: Foundation Reset - Three pillars 본문 작성 (1-2시간) - About 페이지 broad framing 재작성 (3-4시간) - Manifesto 페이지 별도 (2-3시간) - Hero/Sub finalize, 사이트 push (2시간) - Pricing 페이지 초안 (2-3시간) ### Week 2: Bundle MVP Core - Bundle 생성 UI (multi-select) - Bundle 페이지 view (inline preview) - “Copy as context” 기능 - Bundle 편집 (add/remove/reorder) ### Week 3: Universal Polish - Tagging system - Folders - Basic search - Privacy controls (public/unlisted/private) - Bundle UX 다듬기 ### Week 4: Tier 2.5 (Excellence) - Semantic search (embeddings + pgvector) - Bundle versioning (snapshot) - 통합 테스트 ### Week 5: Beta Launch - Beta tester 50명 모집: - Anthropic Discord (MCP) — 15명 - Hacker News 짧은 글 — 15명 - Twitter/X DM — 10명 - Markdown user (Obsidian, HackMD) — 10명 - Onboarding 매끄럽게 - 매주 피드백 콜 ### Week 6: Iteration + Launch Assets - Beta 피드백 fix - Demo 영상 3편 녹화 (Capture/Edit/Use) - Show HN 글 초안 - Manifesto post (1500-2500 단어) - Twitter thread 10 tweets - Product Hunt 페이지 ### Week 7: Public Launch ⭐ - 월: 마지막 점검 - **화: Show HN (오전 9시 PST)** - 수: Twitter thread + dev community - 목: Product Hunt - 금: Manifesto post + AI newsletters ### Week 8: Iteration + Phase 2 시작 - Inbound 처리 - 빠른 fix - 첫 case study - Phase 2 build (Bundle hero campaign 준비) **Launch deadline: Week 7 Tuesday (HARD COMMIT)** --- ## 12. Big Launch Strategy ### Two-track messaging **Track 1 (broad)**: HN/PH/마케팅 - “The fastest way to publish markdown for AI era” - Universal appeal - 즉시 try 유도 **Track 2 (deep)**: Manifesto post + dev community - “Own your AI memory” 철학 - Bundle 차별화 - Movement 동참 유도 ### Manifesto Post **Title 후보**: - “Own your markdown. The case for memory you author.” - “Why I built mdfy: a markdown layer for the AI era” - “Memory is something you write, not something extracted” **Content 구조**: 1. Personal story (왜 만들었나) 2. Problem (현재 AI memory의 한계) 3. Philosophy (5 belief) 4. Solution (mdfy의 architecture) 5. Open invitation (community 동참) **길이**: 1500-2500 단어 ### Open Spec 발표 **“mdfy Bundle Spec v1.0”** - GitHub 공개 - Markdown bundle 표준 제안 - 다른 도구도 implement 가능 - 표준 차지 시도 ### Founder Narrative > “I built mdfy in one month. Multi-surface, MCP-ready, Rust engine. Now going full-time. Why? Because markdown URLs are the right primitive for AI-era knowledge.” HN audience가 좋아하는 builder story. ### Strategic Conversations Launch 전 outreach: - Anthropic (MCP integration) - Cursor (memory layer) - Continue (open source AI coding) - Replit (agent platform) Launch day partnership 발표 가능하면 강력. --- ## 13. KPI 목표 ### Week 8 (Launch) - 1,500+ signup (broad framing 효과) - 300+ Chrome ext install - 75+ paid (Pro $9 → $675 MRR) - 50+ Bundle 생성 (자연 발견) - HN top 5 - 5+ media mentions ### Month 3 - 사용자 8,000+ - DAU 500+ - $5-8K MRR - 1,000+ Bundle - 첫 Build tier 사용자 30명 - Phase 2 (Bundle hero campaign) launch ### Month 6 - 사용자 40,000+ - DAU 2,000+ - $20-30K MRR - 8,000+ Bundle - AI agent integrations 15+ - 첫 partnership 발표 ### Month 9 - 사용자 120,000+ - $50-80K MRR - AI agent integrations 30+ - Team workspace 50+ - Bundle Spec v1.0 release ### Month 12 - $100-150K MRR - 첫 enterprise pilot - “Memory layer” 시장 인지도 top 5 - Strategic conversation 시작 --- ## 14. 차별화 메시지 (정리) ### 강한 메시지 (사용) ✅ **“Own your markdown. Use it anywhere.”** — Hero (universal) ✅ **“Capture, edit, deploy”** — Workflow ✅ **“Owned, edited, portable”** — 3단어 manifesto ✅ **“Your AI memory, authored by you”** — Manifesto/deep audience ✅ **“Bundle as context, deploy anywhere”** — Bundle 차별화 ### 약한 메시지 (피함) ❌ “Black-box memory” (부정확) ❌ “Vendor lock-in” (부정확) ❌ “Don’t let AI extract you” (부정형) ❌ “Write yourself” (manual labor 인상) ❌ “AI memory layer” (Hero에선 너무 좁음, manifesto에선 OK) ❌ “Markdown Hub” (commodity) --- ## 15. Time Allocation (풀타임 40시간/주) ```text Build (40-50%): 18-20시간/주 Launch prep (15-20%): 6-8시간/주 Marketing / Content (15%): 6시간/주 Customer / Beta (10%): 4시간/주 Strategy / Plan (5%): 2시간/주 Rest / Reflection (10%): 3-4시간/주 (필수) ``` 매주 1일 OFF 강제. --- ## 16. Risk Register ### Risk 1: 거인들 진입 (확률 60%, impact 매우 큼) **Mitigation**: - 8주 launch first-mover advantage - Bundle Spec 빠른 발표 - Multi-LLM agnostic - Bundle은 closed system이 따라잡기 어려움 ### Risk 2: Build scope creep (확률 40%, impact 큼) **Mitigation**: - Tier 1 + 2 + 2.5 외엔 launch 전 build 금지 - Week 7 hard deadline - “이 feature 없으면 launch 못 하는가?” — NO면 Phase 2 ### Risk 3: 첫 50명 beta tester 못 찾음 (확률 20%, impact 큼) **Mitigation**: - 채널 4개 다각화 (Discord, HN, Twitter, Markdown community) - Cold outreach, personal network ### Risk 4: Layer 1과 Layer 2 메시지 confusion (확률 30%, impact 중간) **Mitigation**: - Hero는 universal (broad) - Manifesto는 별도 페이지 (deep) - Layer 2는 사용자가 자연 발견 - 두 layer 혼합 marketing 안 함 ### Risk 5: Solo burnout (확률 40%, impact 매우 큼) **Mitigation**: - 매주 1일 OFF 강제 - Month 3+ 자동화 강화 - Month 6+ 외주 검토 --- ## 17. Mantra 매일 자기 점검: > **“이 feature 없으면 launch 못 하는가?”** > NO → Phase 2 list로 미룸 > > YES → 빠르게 build > **“오늘 launch에 가까워졌는가?”** > NO → 무엇이 막고 있는가 > > YES → 계속 진행 > **“Hero는 universal한가, Manifesto는 deep한가?”** > 두 layer 분리 유지 > **“Bundle은 작동하는가?”** > Bundle이 핵심 차별화. 반드시 작동. > **“이게 movement를 만드는가?”** > Big ambition. 단순 product가 아니라 standard 시도. --- ## 18. Long-term Outcome ### Year 1 - $50-150K MRR - 10,000-30,000 paying users - Layer 1 (markdown tool)에서 인지도 - Layer 2 (memory) early adopters - 10+ AI agent integrations ### Year 2 - $300-600K MRR - mdfy Bundle Spec open standard 지위 - C2PA-style consortium 제안 - 첫 enterprise customers - 외부 자본 검토 가능 시점 ### Year 3 - “Memory layer” 시장 top 3 brand - Acquisition 제안 가능 ($50-300M+) - 또는 Series A 본격 성장 - 또는 sustainable indie ($1M+ MRR) --- ## 19. 최종 결정사항 (확정) | 결정 | 답 | | --- | --- | | Direction | Broad markdown tool, Memory는 자연 진화 | | Hero | “Own your markdown. Use it anywhere.” | | Three Pillars | Capture / Edit / Use | | 핵심 차별화 | Bundle (Phase 2 hero) | | Manifesto | 별도 페이지 (deep) | | Wedge | Universal Markdown Tool | | 진입 surface | Chrome extension | | Launch scope | Tier 1 + 2 + Semantic + Versioning | | Pricing | Pro $9, Build $19 | | Launch ambition | Big — standard-setting movement | | 첫 launch 채널 | HN Show HN (Week 7 화요일) | | 차별화 frame | Universal entry + Memory depth | | 경쟁자 인식 | HackMD/Gist 정면 (Layer 1), Mem0 보완재 (Layer 2) | | Open source | mdcore OSS, mdfy.cc SaaS | | Founder commit | 풀타임 100% | | Launch deadline | Week 7 Tuesday (HARD COMMIT) | --- ## 20. 한 줄로 > mdfy는 AI 시대의 markdown 도구다. > > Capture, edit, deploy. 어디서든. > > Power user는 Bundle로 AI memory를 own한다. > > 8주 안에 launch한다. > > Standard를 시도한다. --- *Last updated: 2026-04-27* *Launch deadline: 2026-06-16 (Week 7 Tuesday)* *Next task: Three pillars 본문 작성 (Capture / Edit / Use)* --- ## Summary mdfy.cc is a two-layer markdown tool launching in 8 weeks that aims to be a universal markdown publishing platform for the AI era while evolving into a memory layer where users can author, bundle, and deploy curated knowledge as AI context. The core differentiation is the "Bundle" feature that allows composable, versioned, and deployable collections of markdown documents, positioning mdfy between broad markdown publishing tools (like HackMD) and AI memory extraction services (like Mem0). ## Themes - two-layer product strategy - markdown as AI primitive - user-authored memory bundles - 8-week launch commitment - standard-setting movement ## Key takeaways - mdfy is positioned as a universal markdown publishing tool (Layer 1: capture, edit, share) with optional memory bundling for power users (Layer 2: composable, versioned, deployable context). - Bundle is the core differentiation, defined as composable URL collections that are versioned, deployable to AI agents via MCP/API, and curated by the user, but is deliberately kept as a Phase 2 hero feature during launch. - Launch scope is hard-committed to Tier 1 (tagging, folders, basic search, privacy) plus Tier 2 (bundle MVP) plus Tier 2.5 (semantic search, versioning) within 8 weeks, with Week 7 Tuesday as non-negotiable deadline. - Pricing follows a mass-adoption funnel: Free (unlimited publishing), Pro $9 (private posts, 5 bundles), Build $19 (unlimited bundles, API/MCP), with Team and Scale tiers for future revenue. - The founder commit is full-time solo build with zero marketing and zero users at present, and resource constraints do not exist (no capital or time pressure). - The competitive position is dual-market entry: Layer 1 competes with HackMD and GitHub Gist on AI capture and WYSIWYG editing, while Layer 2 is unique against Mem0/Letta because it is user-authored rather than AI-extracted. ## Insights - The document treats Layer 1 (markdown tool) and Layer 2 (memory bundles) as independently valuable rather than forcing users through a conversion funnel, reducing adoption friction while building toward differentiation. - Bundle versioning and semantic search are positioned as 'Tier 2.5' (excellence features) rather than core launch requirements, allowing launch velocity while preserving long-term moat. - The founder explicitly avoids messaging frames that emphasize negation ('don't let AI extract you') in favor of affirmative positioning ('own your markdown'), suggesting confidence in the product rather than fear-based marketing. ## Open questions / gaps - What specific partnerships or integrations with Cursor, Continue, Replit, or Anthropic are confirmed versus speculative before launch? - How will the semantic search embedding strategy handle cost scaling if the user base grows beyond current projections, and is there a fallback to keyword-only search? - What prevents a large tool (Notion, Obsidian, GitHub) from copying the Bundle feature within 6-12 months, given it is a UX pattern rather than proprietary technology? ## Concepts in this document - **Chrome extension** _(entity)_ Surface within memory.wiki ecosystem. - **Bundle** _(concept)_ A collection of documents and relations in memory.wiki. - **mdfy.cc** _(entity)_ The product being strategically positioned as a two-layer markdown tool with AI capture and memory bundling capabilities. - **Mem0 & Letta** _(entity)_ Competing AI memory solutions that automatically extract memory from conversations rather than user-authored content. - **Own your markdown. Use it anywhere.** _(concept)_ Anchor tagline communicating both ownership and portability as the primary brand positioning. - **Two-layer product architecture** _(concept)_ Core strategic insight: Layer 1 (universal markdown tool) and Layer 2 (power-user memory bundling) that must coexist without forcing memory adoption. - **8-week launch deadline** _(concept)_ Hard commitment to ship by Week 7 Tuesday with defined scope to enforce execution velocity. - **Markdown as AI-era primitive** _(concept)_ Foundational belief that plain-text markdown is the correct format for LLM-native workflows and knowledge portability. - **Capture / Edit / Use** _(concept)_ Three-pillar framework structuring core product value: capturing AI answers, editing in WYSIWYG markdown, and deploying as context. - **Standard-setting movement** _(concept)_ Ambition to establish Bundle spec as open standard (C2PA-style) rather than merely building a product. - **HackMD, GitHub Gist** _(entity)_ Direct Layer 1 competitors in markdown publishing that mdfy aims to displace through AI-native capture. - **User-authored memory** _(concept)_ Philosophical differentiation: intentional user curation rather than AI-extracted memory, positioning against Mem0/Letta approaches. ## Concept relations (within this doc's concepts) - **Chrome extension** captures for **mdfy.cc** - **mdfy.cc** includes **Chrome extension** - **Chrome extension** feeds into **mdfy.cc** - **Two-layer product architecture** layer 2 instantiates **Bundle** - **Own your markdown. Use it anywhere.** hero expresses **Two-layer product architecture** - **Capture / Edit / Use** operationalizes **Own your markdown. Use it anywhere.** - **Bundle** enables spec creation **Standard-setting movement** - **Markdown as AI-era primitive** justifies **Own your markdown. Use it anywhere.** - **Chrome extension** differentiates vs. **HackMD, GitHub Gist** - **Capture / Edit / Use** includes surface **Chrome extension** - **Own your markdown. Use it anywhere.** manifests as **Capture / Edit / Use** _Hub canonical:_ https://memory.wiki/hub/raymindai _Concept digest:_ https://memory.wiki/raw/hub/raymindai?digest=1&compact=1