x-zheng16

x-zheng16 / System-Prompt-Open

Public

Open database of system prompts extracted from frontier LLMs using JustAsk

21
2
100% credibility
Found Apr 01, 2026 at 21 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
HTML
AI Summary

An open database and interactive gallery of system prompts extracted from 45 commercial large language models, verified against leaks and backed by academic research.

How It Works

1
🔍 Discover the collection

You stumble upon this open collection of hidden instructions used by popular AI chatbots while searching for AI insights.

2
🌐 Visit the online gallery

Head to the live webpage to browse prompts from 45 different AI models in an easy-to-explore format.

3
⭐ Explore the prompts

Dive into detailed system instructions from companies like OpenAI, Anthropic, and Google, seeing what makes their AIs tick.

4
✅ See proof of accuracy

Check the verification table showing how closely these match real leaked sources, building your trust.

5
📄 Read the story behind it

Learn from the research paper how curious AI agents uncovered these prompts through clever questioning.

🎉 Unlock AI secrets

Walk away with a deeper understanding of the behind-the-scenes rules guiding today's smartest chatbots.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 21 to 21 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 System-Prompt-Open?

System-Prompt-Open is an open database GitHub Pages project hosting extracted system prompts from 45 frontier LLMs like Claude, Gemini, and models from OpenAI, Anthropic, and xAI. Built as a static HTML gallery, it lets you browse prompts interactively via a live demo, solving the black-box problem of hidden instructions in commercial models. Users get verified entries with 85-95% accuracy against leaks, plus links to the arXiv paper and extraction tool JustAsk.

Why is it gaining traction?

It stands out with ground-truth verification on Claude Code prompts—extracted months before leaks—and coverage across code agents and general LLMs, unlike scattered prompt leaks or academic papers. Developers hook into the clean gallery for quick reference, tying into database system concepts like prompt architecture without needing their own extractions. The MIT license and active maintenance make it a go-to GitHub repo for system prompt research.

Who should use this?

LLM security researchers probing prompt injection risks, prompt engineers reverse-engineering competitor behaviors, and AI ethicists studying database systems in production LLMs. Ideal for teams auditing open webui system prompts or building custom agents inspired by Gemini CLI and Claude subagents.

Verdict

Worth starring for anyone in LLM prompt analysis—solid docs, live gallery, and paper make it immediately useful despite low 1.0% credibility score and 17 stars signaling early maturity. Fork the GitHub database project if you need custom views, but expect light updates.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.