SideKick & DeepWiki + CC,Cline,Roo: This Simple Tool Fixes ALL My Context Issues!

Key Concepts:

• Sidekick: An open-source tool that generates markdown context files for coding agents.
• DeepWiki: A backend system used by Sidekick to create documentation and an AI chat system for understanding repositories.
• MCP (Model Configuration Protocol): A protocol used for AI coders to gather context about a repository.
• Context 7: A tool that uses pattern matching to provide context to AI coders.
• Byte Rover: A memory layer that connects as an MCP in AI coders, allowing them to create and retrieve memories.
• Microsass Fast: A Next.js boilerplate for launching Micros or AI side projects.

1. Introduction to Sidekick

• Sidekick is presented as a simple yet effective open-source tool designed to enhance the performance of coding agents.
• It is available as a hosted service (free) and can also be run locally.
• The primary function of Sidekick is to automatically generate high-quality markdown context files.

2. How Sidekick Works

• Sidekick leverages DeepWiki by Devon in the backend.
• DeepWiki creates documentation and an AI chat system to understand any repository.
• Sidekick works with any public GitHub repository, similar to Context 7 or Git MCP.
• Instead of pattern matching (like Context 7), Sidekick uses an AI model to summarize and generate example snippets.
• It creates markdown files containing information about the repository's functions, tech stack, backend routes, and front-end routes.

3. Comparison with Git MCP and Context 7

• Git MCP is mentioned as a similar tool that indexes GitHub repositories and provides an MCP config for coders.
• Sidekick takes a similar approach but uses markdown and context files instead of MCP.
• This approach aims to provide AI coders with immediate context, eliminating the need to run an MCP first.

4. Local Hosting and Privacy

• Sidekick can be hosted locally for use with private repositories, enhancing privacy.
• The hosted demo may experience rate limits due to DeepWiki.
• Local hosting is free and does not require API keys.

5. DeepWiki MCP as the Backend

• Sidekick uses the DeepWiki MCP as the backend to generate markdown files.
• DeepWiki MCP uses an LLM (Large Language Model) for AI chat, which is leveraged to create the markdown files.

6. Setting Up and Using Sidekick

• The setup process involves cloning the Sidekick repository, navigating to the folder, and running the start dev server file.
• Users can input the link to a GitHub repository and select the coder for which they want rules.
• Sidekick generates a markdown file that can be adapted for various coders.
• The generated markdown file provides detailed information without wasting tokens.
• Instructions are provided on where to place the file for optimal performance.
• Performance improvements of 10-30% are expected, especially for larger repositories.

7. DeepWiki MCP and Byte Rover

• DeepWiki MCP is highlighted for its speed and use of an LLM to summarize repository documentation based on prompts.
• Unlike Context 7, which searches for matching patterns, DeepWiki generates example snippets even if they don't exist.
• Byte Rover is introduced as a memory layer that can connect as an MCP, allowing AI coders to create and retrieve memories.
• Byte Rover can be used with DeepWiki to save queries and access them in memory, avoiding repeated MCP runs.

8. Sponsor: Microsass Fast

• Microsass Fast is a Next.js boilerplate designed to help launch Micros or AI side projects quickly.
• It includes Clerk, Stripe, Resend, HostGSQL, and AI instructions.
• It claims to reduce hallucinations by 90% for vibe coding and offers easy backend integration with Python, Node, and Go.
• Microsass Fast is said to save 50+ hours of setup time.

9. Limitations and Conclusion

• A limitation of Sidekick is that it does not support private repositories in its hosted version.
• Sidekick is recommended as an external tool for summarizing and generating rules for public libraries.
• It is noted to increase performance for AI coders without requiring extensive MCP configuration.

Key Quotes:

• "Sidekick is an open-source tool that is also hosted for free, allowing you to use it even without running it yourself."
• "DeepWiki basically creates documentation and an AI chat system in order for you to understand any repo."
• "This approach makes sure that your AI coder has the required context from the get-go rather than asking it to run an MCP and then gather the info on its own, which is pretty good and more snappy."
• "It is free even if you host locally and won't require any API keys or anything."
• "The bigger your repo is, the better performance you'll get."

Technical Terms:

• Open-source: Software with publicly available source code that can be modified and distributed.
• Markdown: A lightweight markup language used for formatting text.
• Coding Agents: AI-powered tools that assist in software development.
• GitHub Repo: A repository hosted on GitHub, a platform for version control and collaboration.
• Tech Stack: The combination of technologies used to build and run an application.
• Backend Routes: The server-side endpoints that handle requests and responses.
• Frontend Routes: The client-side routes that define the user interface and navigation.
• LLM (Large Language Model): A type of AI model trained on vast amounts of text data.
• API Keys: Unique identifiers used to authenticate requests to an API.
• Next.js: A React framework for building web applications.
• Boilerplate: A pre-built set of code and configurations used as a starting point for a project.

Synthesis/Conclusion:

Sidekick is a valuable tool for enhancing the performance of AI coding agents by providing them with readily available context through automatically generated markdown files. By leveraging DeepWiki's AI-powered documentation and chat system, Sidekick offers a faster and more efficient alternative to traditional MCP-based approaches. While the hosted version has limitations regarding private repositories and potential rate limits, the option to host it locally provides a viable solution for privacy-focused users. The integration with Byte Rover further enhances its capabilities by enabling memory creation and retrieval, making it a comprehensive solution for improving AI coder performance.