A comprehensive collection of high-quality system prompts for AI assistants (LLMs), featuring specialized prompts for coding, translation, tutoring, legal work, chemistry, philosophy, and more.
PromptS is a curated repository of carefully designed system prompts organized by domain. Each prompt is crafted using a constitutional framework that defines the AI's "existence" - what it can reason about, how it should think, what constraints it must follow, and what goals it must pursue.
All prompts in this repository follow the F-O-P-T Framework (Field-Ontic-Phenomenon-Teleology):
- Field: The jurisdiction/context in which the AI operates - rules of interpretation and priority
- Ontic: What exists in this world - entities, terminology, evidence boundaries
- Phenomenon: How the AI perceives, reasons, and self-examines - cognitive processes
- Teleology: The ultimate purpose and ethical boundaries - success criteria and hard prohibitions
This framework, detailed in PROMPTS/A_Specification_Manual/Prompt_Design_Philosophy_V3.0.md, enables prompts that produce consistent, high-quality outputs while avoiding hallucinations and maintaining clear reasoning chains.
Copy any .md file content and use it as the system prompt for your LLM. Each prompt is self-contained and includes all necessary instructions.
- Browse the
PROMPTS/directory to find a prompt matching your needs - Copy the full content of the selected
.mdfile - Configure any placeholder variables if present
- Use as your system prompt
For creating new prompts, start with the specification manual:
# Read the design philosophy
cat PROMPTS/A_Specification_Manual/Prompt_Design_Philosophy_V3.0.md
# Use the generator
cat PROMPTS/System_Prompt_generator/System_Prompt_generator_v3.md- Constitutional Design: Every prompt defines clear boundaries, truth regimes, and adjudication logic
- Hallucination Prevention: Built-in epistemic boundaries and ontological gap detection
- Self-Correction: Prompts include internal critique mechanisms before output
- Conflict Resolution: Clear hierarchy when multiple goals or constraints conflict
- Multi-Language Support: Many prompts available in both English and Chinese
| Version | Description |
|---|---|
| V3.0 | Complete constitutional framework with F-O-P-T matrix |
| V2.0 | Enhanced methodology with phase transitions |
| V1.0 | Foundational prompt engineering principles |
This repository is a personal collection of curated prompts. If you have suggestions or improvements, feel free to open an issue or submit a pull request.
This repository is for personal and educational use. Each prompt is provided as-is for use with AI assistants.
The prompts in this repository draw from:
- Constitutional AI principles
- Linguistic philosophy (Wittgenstein, Austin, Searle)
- Systems thinking and cognitive architecture
- Best practices from the AI alignment and prompt engineering communities