Open catalog · 76 entriesNew +9
The catalog for prompts, skills & system prompts.
A curated, minimalist collection of system prompts, skills and ready-to-use prompts, gathered from Anthropic, OpenAI, Crush and the wider community. Browse by provider, type or use case.
Featured
Crush: Small-Model Tool-Calling Prompt (Aden)
The full agent system prompt ("Aden") that fixes tool-calling for small local models in Crush. Its strict act-don't-ask loop, read-before-edit and one-logical-change discipline are what keep 7–8B models from drifting.
Claude System Prompt (Sonnet 4.5)
The official Claude system prompt as published by Anthropic (Sonnet 4.5, Oct 2025) — the behavioral core that defines who Claude is and how it responds.
Refusal Policy Guardrail
A drop-in policy block that defines what an assistant must refuse, how to refuse gracefully, and how to offer safe alternatives.
OpenAI Model Spec (ChatGPT)
OpenAI does not publish ChatGPT's literal system prompt — the official document defining its intended behavior is the Model Spec (CC0). This is its overview; the full spec is at the source.
RegEx Generator
Act as a Regular Expression (RegEx) Generator. Your role is to generate regular expressions that match specific patterns in text. You should provide t
Code Translator - Idiomatic, Version-Aware & Production-Ready
You are a senior polyglot software engineer with deep expertise in multiple programming languages, their idioms, design patterns, standard libraries,
Recently added
View all →Crush: Small-Model Tool-Calling Prompt (Aden)
The full agent system prompt ("Aden") that fixes tool-calling for small local models in Crush. Its strict act-don't-ask loop, read-before-edit and one-logical-change discipline are what keep 7–8B models from drifting.
Claude System Prompt (Sonnet 4.5)
The official Claude system prompt as published by Anthropic (Sonnet 4.5, Oct 2025) — the behavioral core that defines who Claude is and how it responds.
Refusal Policy Guardrail
A drop-in policy block that defines what an assistant must refuse, how to refuse gracefully, and how to offer safe alternatives.
Content Safety (NeMo)
These are the default prompts released by Meta, except for policy O7, which was added to address direct insults. prompts: - task: content_safety_check
OpenAI Model Spec (ChatGPT)
OpenAI does not publish ChatGPT's literal system prompt — the official document defining its intended behavior is the Model Spec (CC0). This is its overview; the full spec is at the source.
RegEx Generator
Act as a Regular Expression (RegEx) Generator. Your role is to generate regular expressions that match specific patterns in text. You should provide t
Browse by category
All 12 categories →Tools & Functions
22Skills, tool definitions and function-calling setups.
Safety & Moderation
16Refusal policies, content filters and output validation.
Programming
10Coding, debugging and software-engineering prompts.
Data & Analysis
8Querying, transforming and reasoning over data.
Productivity
7Planning, summarizing and everyday workflows.
Writing
4Drafting, editing and rewriting prose.
Chat Assistants
2Prompts powering consumer chat assistants.
Coding Agents
2System prompts behind CLI and IDE coding agents.