Claude Full system prompt leak...... should mine look like this?

Apparently the Claude system prompt is leaked system_prompts_leaks/Anthropic/claude-3.7-full-system-message-with-all-tools.md at main · asgeirtj/system_prompts_leaks · GitHub

Questions:

  1. Studying this….. should our custom Pickaxes look similar?
  2. When we use Claude through API, does it still go through this system prompt? Does our Pickaxe prompt combine with this?

@dougfresh I would say that is part of the AI and automation hype marketing out there. Such a “leak” is highly unlikely, but it is a good hype/viral marketing tactic that some AI guru can use to garner attention within their funnel - other marketers can use it to get attention to their content too.

Have fun. It is an interesting read! GitHub - x1xhlol/system-prompts-and-models-of-ai-tools: FULL v0, Cursor, Manus, Augment Code, Same.dev, Lovable, Devin, Replit Agent, Windsurf Agent, VSCode Agent, Dia Browser, Xcode, Trae AI, Cluely & Orchids.app (And other Open Sourced) System Prompts, Tools & AI Models.

Fun project… I took all these system “leaks” out there and loaded them into a pickaxe and I use that as a foundation to write my system prompts for my other pickaxes. It’s made my system prompts a lot stronger across the board!

I too have “PickAxe builder” Pickaxe and would like to do this as well. Did you put all of the prompts into 1 doc and upload to the Knowledge Base? Individual docs? OR right into the sys prompt?
@Ned.Malki what would you recommend regarding this?

1 Like

Hi @dougfresh the system prompt belongs in the ‘Role’ section of your Pickaxe (see below):

Your IP documents should go in the knowledge base, and your system prompt (think of it as SOPs for your chatbot) should go in the Role section as depicted in the screenshot above.