kju4q/ai-weekend-builds

AI projects to build in a weekend. For developers: requires Python or Node.js, an Anthropic API key, and comfort with the terminal. Starter code and READMEs for each project.

What it is

AI Weekend Builds is a curated set of AI projects designed to be completed in one weekend. Each project comes with a full README, starter code, and the exact prompts the author recommends using. Projects are sorted by difficulty, so you can pick one that matches your skill level and start building immediately.

What you get with each project

Every project folder contains the same four things:

  • README.md — step-by-step setup instructions
  • Starter code — runnable code to get you going immediately
  • Exact prompts — the specific prompts the author used, ready to copy
  • Next steps — ideas for extending the project if you want to go deeper

The four volumes

The repo is organized into four volumes, each with 5 projects. The underlying AI model varies by volume — Volume 4 is built with Codex and uses an OpenAI API key where needed, while Volumes 1–3 use an Anthropic key. Each volume's guide states exactly which key it needs.

Vol 4 — Builds On Your Own Data (August 2026)

Projects that make your own personal data work for you:

  1. Past-You Searcher (Medium, 4–6h) — Everything you've ever written becomes one searchable memory you can ask questions.
  2. Pattern Mirror (Easy–Medium, 3–5h) — Your camera roll becomes a visual representation of the shape of your life.
  3. Lookout (Medium, 4–6h) — Watches your chosen sources and only speaks up when something changes, so you can stop refreshing pages.
  4. Personal API (Medium–Advanced, 5–8h) — A local server where your life's data has addresses.
  5. Freedom Calculator (Advanced, Full day) — Runs 10,000 simulated versions of your next year based on your own assumptions.

Vol 3 — Life-First Local Builds (July 2026)

Tools you build and own, focused on real life rather than work workflows, and local-first so your data stays on your machine:

  1. Life Admin Brain (Medium, 4–6h) — Drop in your lease and policies, then ask them anything — locally.
  2. Money Map (Easy–Medium, 3–5h) — Turns a bank CSV export into a clear view of where your money went, including subscriptions you forgot.
  3. Offline Voice Journal (Medium, 4–6h) — Talk for 2 minutes, get a local transcript with weekly patterns — no cloud involved.
  4. Ambient Life Dashboard (Medium, 5–8h) — Turns an old tablet into a calm, quietly refreshing scene of your day.
  5. Decision Simulator (Advanced, Full day) — Takes a big decision and produces a chart of 10,000 simulated futures.

Vol 2 — 5 New AI Weekend Builds (May 2026)

Projects focused on what's actually shippable now:

  1. Screenshot to Code (Easy, 2–4h) — Photo of any UI becomes a working React component.
  2. AI Daily Digest (Easy, 3–5h) — 5 sources become one morning email of what matters.
  3. Auto-Skill Builder (Medium, 4–6h) — Corrections made in a chat become an installable skill file.
  4. Content Repurposer (Medium, 4–6h) — One post becomes 5 platform-native versions.
  5. AI Inbox Triage (Advanced, Full day) — Connect your inbox and wake up to 3 draft replies ready.

Vol 1 — The original 5

  1. Excalidraw MCP Diagram Agent (Easy, 1–3h) — Describe any system, get a beautiful editable diagram.
  2. One-Command Web Researcher (Easy, 2–4h) — Type a topic, get a full research report from any website.
  3. Personal RAG Assistant (Medium, 5–8h) — Chat with your own docs, PDFs, and notes, with memory.
  4. Multi-Agent Research Crew (Medium, 6–9h) — A team of agents that research, summarize, and report together.
  5. Autonomous Coding Agent (Advanced, Full day) — An agent that reads GitHub issues and writes the code to fix them.

Who it's for

The author states this is for developers comfortable with Python or Node.js and running commands in the terminal. If you're newer to coding, the easy-mode projects are the most approachable starting point.

Author

Made by @qendresahhoti on Instagram.

Related

  • Project
    jingkaihe/matchlock

    Matchlock is an experimental CLI/SDK tool that runs AI agents inside fast-booting, disposable microVMs with network allowlisting, in-flight secret injection (real keys never enter the VM), no host filesystem sharing, and explicit file copy in/out. It supports Linux (KVM) and Apple Silicon macOS, offers Go/Python/TypeScript SDKs, and is MIT-licensed.

  • Project
    ZeroPointRepo/awesome-hermes-skills

    A curated list of add-on skills, plugins, and tools for the Hermes AI agent, with 82 built-in skills, 117 optional ones, and 169 community items—all installable via simple commands.

  • Project
    s3rgeym/hh-applicant-tool

    **Key capabilities** - **AI‑driven**: Uses ChatGPT, Claude, or any OpenAI‑compatible API to write cover letters, answer test questions, and chat responses. - **Resume management**: Auto‑refresh resumes, and filter out unwanted vacancies (e.g., “courier” or taxi jobs) using regex or AI. - **Captcha solving**: Recognizes and auto‑fills captcha. - **Email outreach**: Can also send extra emails to employers. - **Chat auto‑responder**: Replies to employer chats, including handling multiple‑choice buttons. - **Multi‑account**: Supports separate profiles for different hh.ru accounts. - **Local & private**: All data stays on your device; it doesn’t use a headless browser, so it’s lightweight and works on low‑cost servers. - **Fully configurable**: All prompts and filters are editable in a config file.

  • Project
    BruceLanLan/augur

    Augur is a local-first research memory system that helps investors track financial analysis, manage evidence, and compare expert viewpoints over time to ensure research is reproducible and transparent.