Update: Due to legal considerations, this project has been renamed from “Repopack” to “Repomix”
Repopack is a new open-source tool I’ve been testing out. It’s designed to help developers work more efficiently with LLM coding assistants. It tackles a common headache for developers: efficiently sharing entire codebases with large language models.
Most LLM models have strict token limits, making it tough to share large codebases or complex project structures. This often leads to a tedious process of manually copying and pasting relevant code snippets, which is not only time-consuming but also prone to errors.
Another issue is the inconsistent formatting across different parts of a project, which can sometimes confuse LLM models. There’s also the ever-present worry of accidentally sharing sensitive information when selecting code to share.
Most importantly, when we only share isolated parts of a project with an LLM assistant, we rob it of the broader context. This can lead to suggestions that don’t quite fit with the overall architecture or project goals.
What is Repopack?
At its core, Repopack is a command-line tool that packages your entire code repository into a single file. This file is formatted in a way that’s easy for LLM models like GPT-4, Claude, or Gemini to process.
➜ trevorlasn.com git:(master) ✗ repopack --include "src"
📦 Repopack v0.1.43
✔ Packing completed successfully!
📈 Top 5 Files by Character Count and Token Count:──────────────────────────────────────────────────────1. src/content/blog/benchmarks-for-node-bun-deno/index.md (65603 chars, 11330 tokens)2. src/content/blog/common-causes-of-memory-leaks-in-javascript/index.md (42041 chars, 11181 tokens)3. src/content/blog/google-journey-from-search-engine-to-tech-giant/index.mdx (24245 chars, 5025 tokens)4. src/content/blog/micro-frontends-what-they-are-and-when-to-use/index.mdx (19897 chars, 5958 tokens)5. src/content/blog/10-essential-terminal-commands-every-developer-should-know/index.md (18327 chars, 4892 tokens)
🔎 Security Check:──────────────────✔ No suspicious files detected.
📊 Pack Summary:──────────────── Total Files: 123 Total Chars: 738930 Total Tokens: 167841 Output: repopack-output.txt Security: ✔ No suspicious files detected
🎉 All Done!Your repository has been successfully packed.Now if I want to share my project with LLM models, I can simply send them the repopack-output.txt file. This way, the LLM model can see the entire project structure and context, making it easier to provide relevant suggestions.
================================================================Repository Structure================================================================src/ components/ ArrowCard.jsx BackToPrev.astro BackToTop.astro Container.astro ContentContainer.astro Footer.astro FormattedDate.astro Head.astro Header.astro InfiniteScrollPosts.jsx Link.astro SearchModal.astro ThemeSwitcher.tsx content/ blog/ explicit-is-better-than-implicit/ index.mdx 10-essential-terminal-commands-every-developer-should-know/ index.md 2020-programming-trend-predictions/ index.md 39-percent-companies-losing-control-of-it-and-security/ index.mdx a-company-is-not-a-family-its-a-sports-team/ index.mdx a-great-product-doesnt-need-marketing/ index.md ageism-in-tech/ index.mdx aggregate-error-in-javascript/ index.md all-you-need-to-know-about-css-in-js/ index.md amazon-rise-to-tech-titan/ index.mdx amazons-no-weasel-words-rule/ index.md astro-capo/ index.mdx attracting-top-engineering-talent/ index.md become-a-web-developer-in-180-days/ index.md being-a-self-taught-developer/ index.mdx benchmarks-for-node-bun-deno/ index.md build-your-army/ index.mdx cloudflare-ai-content-control/ index.mdx code-wins-arguments/ index.md common-causes-of-memory-leaks-in-javascript/ index.md conways-Law/ index.mdx csp-headers-astro/ index.mdx culture-happens-outside-management/ index.md demystifying-react-hooks/ index.md dependency-time-machine/ index.md docker-with-react/ index.md easy-guide-for-webpack-2-0-from-scratch/ index.md embrace-early-returns-and-intermediate-variables-for-readable-code/ index.md engineering-managers-should-write-code/ index.md eslint-plugin-depend/ index.md evolve-or-become-irrelevant/ index.md frontend-security-checklist/ index.md google-chrome-built-in-gemini-nano/ index.md google-is-killing-information-economics-on-the-internet/ index.md google-journey-from-search-engine-to-tech-giant/ index.mdx how-to-fetch-data-from-an-api-with-react-hooks/ index.md how-to-launch-software-projects-on-time-and-on-budget/ index.mdx how-to-restore-your-passion-for-programming/ index.md how-to-use-redux-with-react-hooks/ index.md increase-react-redux-application-performance-with-reselect-library/ index.md internal-mobility/ index.mdx invisible-columns-in-sql/ index.md is-void-zero-a-threat-to-open-source/ index.mdx its-more-fun-to-be-competent/ index.md lazy-loading-iframes/ index.md make-it-work-first-before-you-optimize/ index.mdx mental-toughness-is-the-best-quality-a-developer-can-have/ index.md mermaid-create-charts-and-diagrams-with-markdown/ index.md micro-frontends-what-they-are-and-when-to-use/ index.mdx minimum-viable-documentation/ index.mdx next-js-react-server-side-rendering-done-right/ index.md objective-c-is-the-ugliest-programming-language-and-a-total-abomination/ index.md open-dyslexic-font/ index.mdx outdated-docs-are-tech-debt/ index.md peaks-js-interact-with-audio-waveforms/ index.md preconnect-to-required-origins/ index.md react-lazy-loading/ index.md react-testing-mock-service-worker/ index.md repopack/ index.mdx setImmediate-vs-setTimeout-in-javascript/ index.md sharp-high-performance-node-js-image-processing-library/ index.md small-habits-big-impact/ index.mdx software-engineer-titles-have-almost-lost-all-their-meaning/ index.mdx specialist-vs-generalist-choosing-your-career-path/ index.mdx speculation-rules-api/ index.md start-with-the-bigger-picture/ index.mdx staying-motivated-while-building/ index.md take-your-writing-seriously/ index.md technical-debt-is-killing-your-business/ index.md the-art-of-effective-onboarding/ index.mdx the-barnacle-strategy/ index.mdx the-credit-vacuum/ index.mdx the-crutch-effect/ index.md the-internet-is-becoming-an-ocean-of-LLM-generated-junk/ index.md the-only-javascript-feature-that-was-deprecated/ index.md the-real-cost-of-meetings/ index.md the-secret-to-being-a-top-developer-is-building-things/ index.md the-what-why-and-how-of-using-a-skeleton-loading-screen/ index.md tips-for-reducing-cyclomatic-complexity/ index.md understanding-javascript-closures/ index.md understanding-vue-suspense/ index.md unrealistic-deadlines-in-software-engineering/ index.md users-can-be-fired/ index.md week-of-coding-can-save-you-hours-of-planning/ index.md what-does-an-entry-level-programmer-need-to-know-exactly/ index.md what-made-apple-great/ index.mdx what-makes-mrbeast-so-successful/ index.md whats-holding-you-back/ index.mdx whats-new-in-express-5/ index.mdx when-regex-goes-wrong/ index.md when-should-you-actually-worry-about-tech-debt/ index.md write-documentation-like-a-journalist/ index.md your-repo-is-a-leaky-ship-probably/ index.md config.ts layouts/ PageLayout.astro lib/ utils.ts pages/ blog/ [...slug].astro otto/ index.astro topics/ [topic].astro index.astro 404.astro index.astro robots.txt.ts rss.xml.ts styles/ global.css utils/ filterPublishedPosts.ts isDraftPost.ts consts.ts env.d.ts types.ts
================================================================Repository Files================================================================
================File: src/components/ArrowCard.jsx================import React from 'react';import { readingTime } from "@lib/utils";
const ArrowCardReact = ({ entry }) => { {* yeah not sharing this, haha *}Think of Repopack like a tar designed specifically for feeding codebases to LLMs. It’s tailored for the unique requirements of working with large language models in a coding context.
Sharing the repopack-output.txt with large language models
Once you’ve generated your repopack-output.txt file, you can easily share it with various AI coding assistants. Here’s how it looks when used with two popular platforms:
Claude
ChatGPT
As you can see, both Claude and ChatGPT can easily process the repopack-output.txt file, allowing them to understand your entire codebase context when providing assistance.
Basic Usage
Getting started with Repopack is straightforward:
npm install -g repopackRun it in your project directory:
repopackTo pack specific files or directories using glob patterns:
repopack --include "src/**/*.ts,**/*.md"Finally, find the repopack-output.txt file in your current directory.
Key Features
- Simplicity: It’s a one-command operation to package your repo.
- AI-Optimized: The output is formatted for easy consumption by AI models.
- Token Awareness: It provides token counts, helping you stay within AI model limits.
- Customizable: You can configure what to include or exclude.
- Security-Minded: It respects
.gitignorefiles and uses Secretlint to avoid exposing sensitive info.
A Word of Caution
While Repopack is useful, it’s important to use it thoughtfully:
- Security: Always double-check that sensitive information isn’t included in the output.
- AI Limitations: Remember that AI tools, while powerful, aren’t infallible. Use their suggestions as a starting point, not gospel.
- Context Matters: Sometimes, less is more. Consider if the AI really needs your entire codebase or just specific parts.
Use it wisely, always be mindful of security concerns, and don’t forget that your expertise and judgment are still the most valuable assets in any development project.