A structured product hub powered by the AISEF framework (v1.2.1).
This repository is the single source of truth for your product's definitions, requirements, architecture, work orders, sprints, and decisions. It follows AISEF (AI Integrated Software Engineering Framework) — a methodology for defining, designing, and tracking products with AI assistance.
Open this repo in Claude Code and say:
> Set up this hub for my product
The aisef-setup skill will walk you through naming your product, choosing a deployment mode, and configuring the hub. Takes about 5 minutes.
- Update
.claude/product.mdwith your product name, description, and team context - Update the
[Product Name]placeholder inCLAUDE.md - Start defining requirements: say "Let's define requirements for this project"
For a full walkthrough, see START-GUIDE.md.
[product]-hub/
├── 01-requirements/ What we're building and why
│ ├── 01-product/ Product overview documents
│ ├── 02-features/ Feature Requirements Documents (FRDs)
│ ├── 03-technical/ Technical constraints
│ └── 04-roadmap/ Feature roadmap
├── 02-blueprints/ How it's designed
│ ├── foundations/ Cross-cutting patterns
│ ├── system-diagrams/ Visual architecture
│ └── features/ Feature-level specs
├── 03-work-orders/ Implementable tasks
│ ├── intake/ New WOs awaiting triage
│ ├── working/ Active WOs
│ └── done/ Completed WOs
├── 04-sprints/ Sprint execution tracking
├── 05-decisions/ Approved deviations and ADRs
├── 06-qa/ Questions and answers
├── 07-issues/ Issues, bugs, signals, observations
├── 08-artifacts/ Reference materials
├── 09-team/ Team roster, dashboard, standups
├── 10-change-requests/ Framework improvements
├── 11-learnings/ Implementation insights
├── .claude/ AI agent configuration and skills
├── .agents/ Gemini workflow adapters
├── .codex/ Codex workflow adapters
└── .github/ CI workflows and agentic automation
This hub works with multiple AI coding agents:
- Claude Code — Primary. 25+ skills in
.claude/skills/. - Codex (OpenAI) — Workflow adapters in
.codex/. SeeCODEX.md. - Gemini (Google) — Workflow adapters in
.agents/workflows/.
All agents share the same canonical skill definitions. Updates to .claude/skills/ are automatically picked up by all adapters.
| Step | What You Say | What Happens |
|---|---|---|
| 1. Define | "Let's define requirements" | Structured interview → product overview documents |
| 2. Specify | "Create an FRD for [feature]" | Feature requirements with testable acceptance criteria |
| 3. Design | "Create a blueprint for [feature]" | Technical architecture with design rationale |
| 4. Decompose | "Break down [feature] into work orders" | Scoped, implementable tasks |
| 5. Execute | "Plan sprint [N]" | Capacity-aware sprint planning |
| 6. Feedback | "A user reported [issue]" | Categorized feedback → work orders |
This hub is your knowledge management center. Implementation repos (where you write code) reference this hub for context:
your-product-hub/ ← This repo (requirements, blueprints, work orders)
your-product-api/ ← Implementation repo (source code)
your-product-web/ ← Implementation repo (source code)
Engineers in implementation repos tell the AI "implement WO-015" — it reads the work order from this hub, reads the blueprint, and builds within those constraints.
See START-GUIDE.md → "Connecting Engineering Repos" for setup instructions.
| File | Purpose |
|---|---|
CLAUDE.md |
AI agent configuration and routing |
CODEX.md |
Codex agent integration |
.claude/framework.md |
AISEF conventions (shared across repos) |
.claude/product.md |
Product-specific configuration |
START-GUIDE.md |
Step-by-step setup and usage guide |
04-sprints/current-sprint.md |
Active sprint board |
03-work-orders/_backlog.md |
Work order index |
- Explicit over implicit — Every decision documented with rationale
- Document the why — Not just what was decided, but why
- Humans decide, AI accelerates — AI assists with drafting; humans own decisions
- People are authors, not operators — Every commit is a conscious act
- Structure frees creativity — Good organization lets you focus on hard problems