lteawoo

lteawoo / SourceLoop

Public

Research runtime for AI tools, built around NotebookLM.

claude codex gemini notebooklm research
19
11
100% credibility
Found Apr 06, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

SourceLoop automates repeatable AI research using Google's NotebookLM to generate cited answers from sources and archive them as local Markdown files.

How It Works

1
🆕 Start your research project

You create a new folder for your research and set it up with a simple command.

2
📝 Name your topic

Pick a name for what you want to research, like 'AI agents', and add any goals.

3
🌐 Open a private research browser

Launch a special, isolated web browser just for your project to keep things safe and separate.

4
📓 Build your AI notebook

Create a smart notebook, add your notes, links, or files, and connect it to your topic.

5
💭 Plan your questions

Generate a list of smart, focused questions tailored to dig deep into your topic.

6
🔄 Run the research

Watch as the AI asks each question and pulls grounded answers with sources cited.

📁 Get your research archive

Everything saves as easy-to-reuse notes in folders, ready for writing, sharing, or building on.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 19 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is SourceLoop?

SourceLoop is a TypeScript CLI runtime that automates deep research workflows for AI tools like Codex, Claude Code, or Gemini CLI, built around NotebookLM. It manages a dedicated Chrome browser, plans batches of targeted questions, runs grounded Q&A on your sources, and archives answers with citations as local Markdown files. No more vanishing browser tabs or weak prompts—get reusable github research data from urls, files, or transcripts.

Why is it gaining traction?

It skips RAG hassles like vector DBs by leveraging NotebookLM's multi-source reasoning and built-in citations, while adding topic-first planning and full-batch runs via sourceloop cli. Developers dig the Obsidian-friendly outputs for github research blog posts or surveys, plus audit logs for runtime error tracking in research github copilot flows. Stands out for turning ephemeral AI chats into versioned research repos without custom research runtime libraries.

Who should use this?

AI agent builders needing grounded research without hallucinations, like those scripting github research group koch analyses or research paper template generation. Ideal for indie devs doing research on webassembly runtimes a survey, or solo researchers chaining source loop gata only pipelines from NotebookLM. Skip if you're not in deep research tools.

Verdict

Worth a spin for NotebookLM fans—sourceloop cli delivers real research-to-runtime value in 19 stars. At 1.0% credibility, it's early (rough UI selectors, citation fidelity), but solid docs and tests make it playable now; fork for production.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.