stream-coding

🥇Gold

Stream Coding is a revolutionary methodology that enhances AI-accelerated development, enabling developers to boost their coding velocity by 10-20x. It provides essential resources like the official SKILL.md and a comprehensive Manifesto for effective implementation.

8840Updated 1mo ago

Intermediate15 minutes to implementdevelopment

Saves ~240 min per use

Quick InstallView Source

claude install frmoretto/stream-coding

Works with:

ClaudeCursor

Overview

About This Skill

How to Use

1. **Prepare Your Codebase**: Document your current development velocity metrics (e.g., PR review time, build duration) and identify 1-2 pain points to target first. Use tools like GitHub Insights or Linear for baseline data. 2. **Adopt the Manifesto**: Familiarize yourself with the Stream Coding Manifesto's 10 principles (SKILL.md §1.1). Focus on the 'Stream State', 'Live Design', and 'AI Co-Pilot' principles for immediate gains. 3. **Pilot a Toolchain**: Start with one optimization. For example, integrate Rive's real-time API if your team struggles with design-code sync delays. Use Rive's collaboration features to co-develop animations with designers in real time. 4. **Measure and Iterate**: After 2-3 sprints, compare velocity metrics against your baseline. Use the Manifesto's 'Velocity Score' framework (SKILL.md §6.2) to quantify improvements. Double down on what works, and phase out underperforming optimizations. 5. **Scale Gradually**: Expand to other principles (e.g., parallelized state management) once your pilot proves successful. Document lessons learned in a team Wiki to institutionalize Stream Coding practices.

Use Cases

Accelerate software development cycles

Reduce technical debt in projects

Implement specification-driven development practices

Enhance collaboration among developers

Best For

GrowthCS

Setup & Installation

Quick Install

Terminal

claude install frmoretto/stream-coding

Alternative Install (Git Clone)

git clone https://github.com/frmoretto/stream-coding

Requirements

Claude Code or compatible AI agent
Works with: Claude, Cursor

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Act as a Stream Coding expert. Using the [SKILL.md] and the Stream Coding Manifesto as your guide, analyze the following codebase: [CODEBASE_DESCRIPTION_OR_LINK]. Identify bottlenecks in development velocity and suggest 3-5 high-impact Stream Coding optimizations. For each optimization, provide: 1) A specific implementation strategy, 2) Expected velocity gains, and 3) Risks to mitigate. Reference the Manifesto principles directly in your recommendations.

Example Output

I analyzed the open-source e-commerce platform 'ShopFlow' (GitHub: shopflow/ecommerce), a React + Node.js monolith with 150k lines of code. Using Stream Coding principles, I identified three critical bottlenecks:

1. **Parallelized State Management**: The current Redux store architecture serializes all state updates, creating a 400ms delay per user action. Implementing the Manifesto's 'Stream State' pattern (SKILL.md §4.2) with Redux Toolkit's createSlice would parallelize state mutations. Expected gain: 3.2x faster state updates (from 400ms to 125ms). Risk: Race conditions in cart operations. Mitigation: Add optimistic UI updates with conflict detection.

2. **Automated Design-Code Sync**: The team manually recreates Rive animations in React components, causing 2-week delays per animation. Adopting the Manifesto's 'Live Design' principle (SKILL.md §3.1) by integrating Rive's real-time API would auto-generate React hooks from animations. Expected gain: 5x faster UI implementation (from 14 days to 2.8 days). Risk: Animation performance degradation on mobile. Mitigation: Use Rive's runtime optimization flags and test on low-end devices.

3. **AI-Powered Code Reviews**: The current manual PR review process takes 3.5 days per feature. Implementing the Manifesto's 'AI Co-Pilot' principle (SKILL.md §5.3) with GitHub Copilot + custom Stream Coding rules would auto-generate 70% of boilerplate and flag anti-patterns. Expected gain: 4x faster reviews (from 84 hours to 21 hours). Risk: False positives in custom rule enforcement. Mitigation: Start with a 10-rule pilot set and expand based on precision metrics.

**Velocity Impact**: Combining these optimizations could reduce ShopFlow's feature delivery time from 6 weeks to 1.5 weeks—a 4x improvement. The Manifesto emphasizes incremental adoption (SKILL.md §2.4), so I recommend piloting the Rive integration first due to its immediate ROI on UI work.