Documentation engine for human-AI collaborative system design
npm install @futurefoundaihq/foundryspecThe Documentation Engine for AI-Human Collaborative System Design.
FoundrySpec is not just a documentation tool; it is a Living Digital Twin of your system's architectural intent. Designed for the era of AI-assisted engineering, FoundrySpec ensures that your documentation remains the Single Source of Truth (SSoT)βa centralized, authoritative reference point that both humans and AI Agents can trust to avoid data fragmentation and drift.
---
In a world of rapid AI development, traditional documentation fails because it is unstructured, ambiguous, and stale. This leads to AI hallucinations, context fragmentation, and architectural decay.
FoundrySpec solves this by providing:
* π€ AI-First Design: Optimized for LLM context windows. No prose-heavy fluffβjust structured, machine-readable specifications (Mermaid + Markdown).
* πΊοΈ Discovery-First Workflow: Mandates a "Discovery Phase" (Personas, Journeys, Requirements) before a single line of code is written.
* π Zero-Question Traceability: Link high-level requirements directly to code components. Achieve "Zero-Question Implementation" readiness.
* β‘ Zero-Config Auto-Discovery: Simply create a folder in assets/, and the engine handles the rest. No configuration bloat.
* π Hot-Reload Dev Experience: Real-time feedback loop. Edit your diagrams and see the changes instantly in the Documentation Hub.
---
---
bash
npm install -g @futurefoundaihq/foundryspec
`$3
`bash
foundryspec init "My Awesome Project"
cd foundryspec
foundryspec serve
`$3
`bash
foundryspec upgrade
foundryspec build
`$3
The best way to get an AI Agent (like Gemini, Jules, or Claude) started is to simply tell it:
> "Run foundryspec help to understand the system architecture and our design-driven workflow."---
π€ For AI Agents (Jules, Gemini, etc.)
FoundrySpec provides you with a Structured Context Layer. When working in a FoundrySpec-enabled repository:
1. Consult Discovery: Understand the user's intent and personas.
2. Verify Traceability: Check which requirements your task satisfies.
3. Align with Architecture: Ensure your implementation matches the Container and Component diagrams perfectly.
4. Build & Validate: Use
foundryspec build` to check for broken links or syntax errors before committing.---
---
Build the future. Document with Intent. Use FoundrySpec.