Compospec

Stop building onboarding flows from scratch. This AI ready template solves that. Here's a production-ready User Onboarding Context MD. Most onboarding flows are built ad-hoc: - Generic validation rules - Forgotten edge cases   - Missing analytics events - Zero documentation for AI tools This template changes that. 🗂️ What's Inside: → 1 Product Context (high-level strategy) → 3 Modules (signup, setup, activation)   → 9 UI Cards (every screen, every rule, every edge case) Each md file includes: ✓ Validation logic (email typos, password strength, slug generation) ✓ Error handling (network failures, race conditions, rate limits) ✓ Analytics events (track everything from signup to upgrade) ✓ Edge cases (Unicode slugs, disposable emails, time zones) ✓ Testing scenarios (valid/invalid inputs) No prose and ambiguity. Just structured rules AI can execute. Structured for AI: Feed these markdown filesto Claude, ChatGPT, or any AI tool: → Figma mockup with validation states and error messages Build the workspace API: OpenAPI spec with edge cases and rate limits Write React components: → Production-ready TypeScript with form validation Generate test cases: → Complete test suite covering all scenarios No more starting from zero. You can: ✓ Clone to your workspace (customize for your product) ✓ Export as PDF/MD (share with your team)   ✓ Share public links (send specific cards to developers) ✓ Edit every card (it's your template, adapt it freely) Built with Compospec, the AI-native spec tool.  Links in comments: - Free template - Live flow diagram Turn your product ideas into structured context that AI can actually understand.

This is an 86-card structured spec. Sample product: AI Content Moderation Platform. Built in Compospec's 6-level hierarchy. Product → Module → UI → Section → Component → Element. Structured. Semantic. AI-ready. New features: MD export → AI builds without hallucination. Link sharing → No signup, just review. Template Marketplace → One-click import. Spec-Driven Development for AI. 👉 Flow link in comments

You're writing CLAUDE md, memory-bank/, SKILL md by hand. Folder by folder. Every single project. Most teams are already writing this. PRD in Notion/Confluence. Flows in Figjam/Miro. Context in Slack. CLAUDE md, memory-bank/, SKILL md etc. all written by hand, folder by folder. Scattered. Lost. Rewritten every time. Here's what changes when you structure it once in Compospec: - CLAUDE md generates from your Product card - skills/SKILL md comes from your Component + Element cards - memory-bank/ maps to your full 6-level hierarchy - Auto generated flow diagram exports ready for Claude Code - Share any spec publicly, no signup needed One structured spec. Every file Claude Code needs. Ready to drop into .claude/ You're already writing it. Write it once. Write it right. ↓ Live sample flow diagram link in the comments. ↓

Developers think in syntax. But, you speak in paragraphs. Developers structure everything on screen hierarchically, following DIV logic. You just dump your thoughts into plain text. Developers need conditions, validations, and UI specs for every single field. You send a design and hope it speaks for itself. Developers want to see user journeys as diagrams. Let's be honest; most of us never draw them. Developers want to reach the spec of a radio button in seconds. You can't even find it yourself. AI coding tools expect the same things. But you're still burning tokens with prompt bombardment. If you can't deliver a screen like this as a spec, in a semantic "UI grammar" structure, from high level to micro level, with parent-child relational hierarchy; This is "time theft". 👇 Look at this screen. Every element, every section, every component. Semantic, navigable, spec-ready.

Compospec is designed to solve a problem every team faces, turning scattered product information into something organised and usable. For product teams, it’s a structured place to write and manage requirements instead of relying on long documents and disconnected tickets. 🎟️ For developers, it provides clear context so features are built from agreed specs, not assumptions. For designers, it links decisions to components so everyone understands why something exists. For marketers and business stakeholders, it creates a shared language to turn ideas and feedback into actionable product changes. And for leadership, it offers visibility, version control and a reliable source of truth across the entire product. At its core, Compospec gives everyone the same picture, in one connected system, so teams spend less time searching and more time building. 🏗️ #ProductManagement #ProductTeams #Documentation #Requirements #SoftwareDelivery #Compospec

Imagine buying a complex LEGO set with no numbered bags. All the pieces are poured into one giant box and you’re told to start building. Finding a single specific piece becomes a hunt through thousands of others, and progress is slowly than January sales... That’s how most product documentation works today. Requirements, decisions and context are all mixed together with no real structure. Compospec is like having clearly labelled LEGO bags. Every part of the product has its own organised place, so teams always know exactly where to look and exactly what fits where. 🌐 www.compospec.com #ProductManagement #ProductSpecs #Documentation #ProductDelivery #ProductTeams

Most teams don’t struggle to create ideas. They struggle to organise them. Product knowledge usually ends up scattered across documents, tickets and conversations, which means important details get lost and clarity slowly fades over time. Instead of one big, messy document, Compospec breaks requirements into clear, connected layers so every decision has a logical home and every part of the product can be traced back to its purpose. When information is organised properly, teams stop hunting for answers and start building with confidence. Not more documentation. Better documentation. #ProductManagement #ProductSpecs #Requirements #ProductDelivery #StructuredWork #Compospec

Great research is only useful if a delivery team can actually work from it. Too often research ends up as long reports, workshop notes and scattered documents that never quite translate into real product requirements. That’s where projects start to drift. Teams are handed “insights” instead of specifications and asked to fill in the gaps themselves. Turning research into usable specs means taking what you’ve learned and structuring it properly: Breaking findings into clear requirements. Connecting them to features and components. Defining scope, behaviour and constraints. Making sure nothing important gets lost along the way. Compospec bridges that gap. It gives teams a way to convert raw research into structured, linked requirement cards that are clear, actionable and traceable from idea to delivery. Because knowledge only creates value when it becomes something people can build from. #ProductManagement #UserResearch #ProductSpecs #Documentation #ProductDelivery #Compospec

Sixty years ago, COBOL promised plain English code that anyone could write. The dream was business people creating software without programmers, the reality? It still needed structure, organisation and specialists. What’s Different Now: In 2026, AI finally delivers the missing piece. Business analysts can write specs in plain English, and AI can translate them into working code. The original COBOL promise feels fulfilled. But the same problem remains, when something becomes easy to write, it becomes harder to manage. COBOL made code easier, and technical debt exploded. AI makes specs easier, and prompt debt is exploding. More specs, more versions, more scattered context, less clarity. COBOL gave us readable code, Git gave us control over it, AI gives us executable specs, but we don’t yet have proper control over them. We need a real spec management layer: • Semantic structure, not just free text. • Clear links between requirements. • Dependencies and hierarchies. • Version history. • Tracking of what’s live and what’s not. Without that, AI simply accelerates chaos. The Lesson - Making something easy to write doesn’t make it easy to manage. We solved this for code, now we need to solve it for specs. Credit to @Chris Rickard for the COBOL insight that sparked this thinking. #SpecDrivenDevelopment #PromptDebt #AIHistory #ProductManagement

With Compospec, the flow lives in one connected structure from day one. Requirements are broken into clear, hierarchical cards, so every part of the product has a logical home. Decisions stay attached to the exact component they relate to, not buried in chat history. When something changes, you update it once and everyone sees the same source of truth. No hunting through files. No guessing what was agreed. No reinventing work that already happened. Manual processes treat documentation as something you tidy up at the end. Compospec treats documentation as part of the product itself. That’s why teams move faster, argue less and deliver with far more confidence. #ProductManagement #ProductSpecs #Documentation #ProductDelivery #Compospec