You just copied the commands that stop your AI assistant from forgetting your project structure every single session. No more re-explaining your architecture. No more context window limits breaking your flow.

Knowns creates a persistent memory layer that your AI tools can query. It's like giving ChatGPT a photographic memory of your entire codebase, architecture decisions, and undocumented quirks. The 112 developers who starred it on GitHub in the last week are already shipping faster.

TL;DR: Why This Matters Now

• What: Knowns is an open-source memory layer that indexes and stores your project's context for AI assistants.
• Impact: It solves the critical 'context amnesia' problem where AI forgets your codebase between sessions.
• For You: Cut onboarding time for new AI tools from hours to seconds and get consistently accurate code suggestions.

The Goldfish Memory Problem

Every developer using AI tools faces this: you spend 20 minutes explaining your project's architecture to ChatGPT or Claude. You get perfect suggestions. Then you close the tab.

Tomorrow, you're back to square one. The AI has forgotten everything. This context amnesia wastes hours weekly and makes AI suggestions increasingly generic over time.

Knowns fixes this by creating what the developers call "the memory layer for AI-native development." It's not another AI tool—it's the infrastructure that makes your existing tools smarter.

How It Works (The Simple Version)

When you run knowns index, it:

• Scans your entire codebase structure
• Extracts relationships between files and modules
• Creates a searchable knowledge graph of your project
• Stores this persistently in a local vector database

Your AI tools can then query this memory layer before answering questions. Instead of working blind, they have full project context instantly.

The Real-World Impact

Early adopters report 3 immediate benefits:

1. Onboarding new team members (human or AI) takes minutes, not days. The memory layer contains architectural decisions that usually live only in senior developers' heads.

2. Code suggestions become project-aware. Instead of generic "here's how to write a function," you get "here's how to write this function following YOUR project's patterns."

3. Documentation stays current automatically. As your code evolves, Knowns updates its understanding. No more outdated README files misleading everyone.

Getting Started Beyond The Box

The quick-value box gets you running. For maximum impact:

Integrate with your existing workflow: Knowns works with GitHub Copilot, Cursor, and any AI that accepts context. It's infrastructure, not replacement.

Set up incremental indexing: Configure it to watch for changes and update the memory layer automatically. Your AI always knows about yesterday's refactor.

Share team memories: The upcoming enterprise features let teams share a unified project memory. Everyone's AI assistant operates from the same understanding.

The Bigger Picture

Knowns represents a shift in how we think about AI-assisted development. We've been treating AI as disposable consultants who forget everything after each meeting.

This tool treats AI as persistent team members. It's the difference between hiring a contractor for one task versus having a senior engineer who knows your codebase intimately.

The 112 stars in a week signal developers are tired of AI's goldfish memory. They want tools that learn and remember. Knowns delivers exactly that.