by asgeirtj
Extracted system prompts from Anthropic - Claude Fable 5, Opus 4.8, Claude Code, Claude Design. OpenAI - ChatGPT 5.5 Thinking, GPT 5.5 Instant, Codex. Google - Gemini 3.5 Flash, 3.1 Pro, Antigravity. xAI - Grok, Cursor, Copilot, VS Code, Perplexity, and more. Updated regularly.
# Add to your Claude Code skills
git clone https://github.com/asgeirtj/system_prompts_leaksGuides for using ai agents skills like system_prompts_leaks.
Last scanned: 4/15/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-04-15T06:03:11.321Z",
"semgrepRan": false,
"npmAuditRan": true,
"pipAuditRan": true
}system_prompts_leaks is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by asgeirtj. Extracted system prompts from Anthropic - Claude Fable 5, Opus 4.8, Claude Code, Claude Design. OpenAI - ChatGPT 5.5 Thinking, GPT 5.5 Instant, Codex. Google - Gemini 3.5 Flash, 3.1 Pro, Antigravity. xAI - Grok, Cursor, Copilot, VS Code, Perplexity, and more. Updated regularly. It has 47,202 GitHub stars.
Yes. system_prompts_leaks passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/asgeirtj/system_prompts_leaks" and add it to your Claude Code skills directory (see the Installation section above).
system_prompts_leaks is primarily written in JavaScript. It is open-source under asgeirtj on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh system_prompts_leaks against similar tools.
No comments yet. Be the first to share your thoughts!
Based on votes and bookmarks from developers who liked this skill
As seen in The Washington Post: See the hidden rules behind AI. Then use them to rewrite this article. (May 11, 2026)
The purpose of this repo is to document the System Prompt instructions for all the AI chatbots out there - Claude, ChatGPT, Gemini etc.
🆕 Diff: Claude Opus 4.8 → Claude Fable 5 — see exactly what changed in the claude.ai system prompt for Anthropic's newest model
| What | Date | Link |
|---|---|---|
| Claude Design (Opus 4.8 — full prompt + 48 tools + 16 skills + 9 starter sources) | June 26, 2026 | System prompt |
| GitHub Copilot for macOS (app) | June 18, 2026 | System prompt |
| GPT-5.5 Codex (full prompt) | June 18, 2026 | System prompt |
| Claude Fable 5 | June 9, 2026 | System prompt · Diff vs Opus 4.8 |
| Claude Opus 4.8 | June 9, 2026 | System prompt · Official |
| Claude Code Glob & Grep tools | June 9, 2026 | Glob · Grep |
| Claude Code (Opus 4.8) | May 28, 2026 | System prompt |
| Claude Code & Cowork | May 28, 2026 | Claude Code · Cowork · Cowork Dispatch |
| GPT-5.5 | May 24, 2026 | Thinking · Instant · API · Pro API |
| Perplexity Computer | May 21, 2026 | System prompt |
| VS Code Copilot Agent | May 21, 2026 | System prompt |
| Docker Gordon AI | May 21, 2026 | System prompt |
| Gemini 3.5 Flash | May 20, 2026 | System prompt · AI Studio · Tools |
| Antigravity CLI | May 20, 2026 | System prompt |
| Zed AI | May 16, 2026 | System prompt |
| Grok Expert | May 11, 2026 | System prompt |
| Model | Prompt |
|---|---|
| Claude Fable 5 | System prompt |
| Claude Opus 4.8 | System prompt |
| Claude Code (Opus 4.8) | System prompt |
| Claude Opus 4.7 | System prompt |
| Claude Code (Opus 4.6) | System prompt |
| Claude Opus 4.6 | System prompt |
| Claude Sonnet 4.6 | System prompt |
| Claude.ai | Anthropic Reminders |
| Integrations | Cowork · Cowork Dispatch · Desktop Code · Design · Mobile iOS · In Chrome · For Excel · For Word · In PowerPoint · Default Styles |
| Claude Code extras | Glob tool · Grep tool · Deferred tools · Docs assistant · Bundled skills |
Published (claude_behavior at release date, not updated) |
Opus 4.8 · Opus 4.7 · Opus 4.6 · Sonnet 4.6 · All versions |
| Without tools | Opus 4.6 · Sonnet 4.6 |
| Raw prompts | Opus 4.6 · Opus 4.6 (no tools) · Sonnet 4.6 · Sonnet 4.6 (no tools) |
| Visualize | Visualization |
| Opus 4.5 | System prompt |
| Sonnet 4.5 | System prompt |
| Sonnet 4 | System prompt |
| Opus 4.1 Thinking | System prompt |
| Sonnet 3.7 | System prompt · With tools · Full w/ tools · Human-readable |
| Model | Prompt |
|---|---|
| GPT-5.5 | Thinking · Instant · API · Pro API · Codex · Friendly · Pragmatic |
| GPT-5.4 | API · Thinking · Codex · Codex Mini |
| GPT-5.3 | Codex · Spark · Codex API · Chat API · Instant |
| Codex CLI | Per-model prompts · Spark · Plan mode · Personas · Auto-review |
| Tools | Web search · Deep research · Python · Python code · Canvas · Image gen · Memory · Advanced memory · File search |
| Policies | Image safety · Automation context |
| GPT-5.2 | Mini (free) · Thinking · Codex |
| o4-mini | System prompt · High |
| o3 | System prompt |
| ChatGPT Atlas | System prompt |
| GPT-5.1 personalities | Default · Friendly · Professional · Candid · Cynical · Efficient · Nerdy · Quirky |
| GPT-5 | Agent mode · Thinking · Cynic · Listener · Nerdy · Robot · Codex · Codex Mini |
| GPT-4.5 | System prompt |
| GPT-4.1 | Full · Mini |
| GPT-4o | System prompt · WhatsApp · Advanced voice · Legacy voice |
| Monday GPT | System prompt |
| GPT-4o new personality | System prompt |
| Study and learn | System prompt |
| Image safety policies | System prompt |
| API variants | GPT-5 reasoning (high) · o3 high · o3 med · o3 low · o4-mini high · o4-mini med · [o4-mini low](OpenAI/API/o |