DocsAI — Product Requirements Document

One-liner

The publish layer between AI agents and humans. Markdown in → professional branded pages out, with access control and API.

The Problem

Every company using AI agents (Claude Code, Cursor, Copilot) generates markdown files that sit in git repos, invisible to anyone who doesn’t cat files in a terminal. The output is trapped.

For Cofoundy specifically:

• Claude Code writes meeting notes, brand guides, proposals, project docs, deliverables — all as .md files
• Those files live in repos that clients never see
• Today we send PDFs on WhatsApp or links to Google Docs — things get lost, there’s no single source, no branding, no access control
• Team members write CLAUDE.md files that are rich with knowledge, but nobody reads them because raw markdown in a repo is hard to navigate

For the market broadly:

• Every agency, consultancy, and dev shop using AI coding tools has this problem
• The gap between “AI generated a great document” and “client can see it professionally” is manual work: copy to Notion, export to PDF, upload to Drive, send link
• That manual step kills the speed advantage of AI

The Insight

Documents shouldn’t be edited by humans anymore. They should be authored by AI agents and published automatically. The human’s job is to review, approve, and share — not to format, upload, and manage links.

The closest analogy: Vercel for documents. Git push → live branded page with a URL. But instead of code deploying as a web app, markdown deploys as a professional document.

Users & Personas

Priority order: AI Agent > Client reader > Team member > Public visitor

Core Concepts

Document

A markdown file with YAML frontmatter:

---
title: "Brand Guidelines — Acme Corp"
role: client
version: 3
author: andre
created: 2026-03-22
tags: [branding, deliverable]
---

Project Space

Documents are grouped by project folder. Maps 1:1 to Cofoundy’s project structure:

• cofoundy/ — internal docs (team only)
• client-acme/ — client deliverables (client + team)
• blog/ — public content

Access Roles

Technical Architecture (MVP)

graph LR
  CC[Claude Code<br/>/publish skill] -->|git push| REPO[docs-ai repo<br/>Astro + markdown]
  REPO -->|auto-deploy| CF[Cloudflare Pages]
  CF --> CFA{Cloudflare Access}
  CFA -->|team emails| T[Team]
  CFA -->|client emails| CL[Client]
  CFA -->|no auth| P[Public]

V1 adds: API server (Cloudflare Workers) between Claude Code and the repo.

V2 adds: Database (D1/Turso) for multi-tenant, user management, analytics.

MVP Features

1. Astro static site with markdown content collections (GFM, syntax highlighting, tables, Mermaid.js, callouts, auto TOC)
2. Cofoundy branding applied automatically (header, footer, fonts, colors, print-friendly CSS)
3. Frontmatter-based routing (role: public/team/client → Cloudflare Access)
4. Deploy: Cloudflare Pages (git push → auto-deploy, docs.cofoundy.dev/{project}/{slug})
5. /publish skill (copies .md to docs repo with frontmatter, commits + pushes, returns live URL)
6. Index pages (per-project + global)

Validation Framework

Why This Could Be a Startup

1. Timing — AI coding agents are exploding. Every team needs a publish layer.
2. Wedge — Agencies have the sharpest pain (multiple clients, professional output matters).
3. Expansion — Legal, marketing, consulting — anyone using AI to write.
4. Moat — The /publish skill ecosystem. AI agent → distribution channel.
5. Revenue — Free for public docs. $29/mo for client spaces. $99/mo for white-label.