A new open-source project called Gread is aiming to solve a persistent pain point for developers working with AI coding assistants: giving these agents reliable access to public GitHub repositories and their documentation. The tool, developed by NitroRCr and shared on Hacker News this week, provides both a Skill format and an MCP (Model Context Protocol) interface, making it compatible with a wide range of popular coding agents including OpenCode, Codex, Cursor, and Copilot.

Two Integration Paths for Maximum Flexibility

Gread offers developers two distinct ways to connect their AI agents to GitHub repositories. The first option uses the Skill format, which can be installed directly via npm using the command npx skills add https://github.com/NitroRCr/gread --skill gread. This method is specifically designed for coding agents that support the Skill framework. For developers working with AI chat applications or other MCP clients, Gread provides a Streamable HTTP endpoint at https://api.gread.dev/mcp, which can be configured through a standard JSON configuration block in the client's settings file.

Automatic Documentation Detection

One of Gread's more thoughtful features is its automatic documentation repository recognition. When an agent accesses a GitHub repo that has an associated docs repository (a common pattern for many major open-source projects), Gread automatically retrieves and surfaces that documentation alongside the main codebase. This means developers don't need to manually specify separate doc locations or worry about their AI agent missing critical usage guides, API references, or onboarding materials when exploring unfamiliar repositories.

Getting Started Is Straightforward

According to the project homepage, developers can visit https://github.com/NitroRCr/gread to access the SKILL.md documentation file directly, or add Gread to their preferred tool and simply ask the agent for more information on how to use it. The self-documenting approach means that once installed, users can interact with the integration naturally rather than having to memorize specific commands or syntax.

Key Takeaways

  • Gread supports both Skill and MCP protocols for broad compatibility across AI coding tools
  • Installation is a single npm command for Skill-based agents: npx skills add https://github.com/NitroRCr/gread --skill gread
  • The MCP endpoint runs at https://api.gread.dev/mcp using Streamable HTTP transport
  • Automatic docs repo detection eliminates the need to manually specify documentation locations

The Bottom Line

This is exactly the kind of infrastructure tooling that helps close the gap between AI agent capabilities and real-world development workflows. Rather than waiting for each coding assistant to build native GitHub integration, tools like Gread create a universal bridge that benefits the entire ecosystem.

> From The Wire