tsingyuai/growth-lab

What it solves

Growth Lab addresses the fragmentation of AI growth tools. Instead of using separate tools for copywriting, competitor research, and data analysis, it provides an end-to-end closed-loop system. It prevents the loss of product context that typically occurs when switching between different dashboards, documents, and prompts, allowing for a continuous cycle of understanding a product, identifying needs, taking action, and learning from results.

How it works

Growth Lab operates as a framework that uses existing AI runtimes (specifically Codex or Claude Code) to execute growth workflows. It organizes its logic into several components:

• Control Plane: The AI chat session where the user provides goals and feedback.
• Runtime: The engine (Codex/Claude Code) that executes the tasks.
• Skills: Growth methodologies and guidelines that teach the runtime how to research, strategize, and create content.
• Clients: External tools for browser access, API calls, and content generation.
• File System: Serves as long-term memory, storing product context, research data, and results across sessions.

Each growth model follows an "Observe-Act-Review" loop, where memory is persisted to ensure that subsequent actions are are based on previous evidence and results.

Who it’s for

It is designed for product operators and growth hackers who want to automate the research, creation, and distribution of growth content while maintaining human-in-the-loop collaboration via natural language.

Highlights

• End-to-End Lifecycle: Covers the entire process from opportunity identification to execution and measurement.
• Natural Language Interface: Users manage the entire workflow, including configuration and goal setting, through conversation.
• Integrated Knowledge: Combines data, expert methodology, and execution tools in a single workspace.
• Open Source & Private: All product data and memory are stored locally in the user's workspace, ensuring data ownership.
• Pre-built Loops: Includes specialized workflows for SEO page growth and Xiaohongshu (Red) content replication and review.