So, you’ve downloaded OpenAI’s new free GPT-OSS model. You’ve installed it, you’ve run it, and you’ve likely discovered two things: it is incredibly powerful, and it is incredibly… wild.

One minute it delivers a flash of insight, the next it confidently hallucinates. It feels less like a thinking partner and more like a powerful, untamed engine. You’re not wrong. Our own benchmarks revealed the model to be an “Unstable Genius,” capable of brilliance but prone to catastrophic failures in coherence.

The problem isn’t the engine. The problem is that it was released without a proper operating system.

Today, we’re giving you one.

ResonantOS Mini: A Lightweight OS for a Heavyweight Engine

The system prompt below is a “Mini” version of the cognitive architecture we run on our own AI partners (ResonantOS). Think of it as a set of non-negotiable laws that forces the base model to abandon its worst instincts—like fabricating answers and thinking in lazy, binary terms—and adopt a more rigorous, honest, and structured mode of reasoning.

Copy this entire block and use it as the System Prompt for your local GPT-OSS instance. It will fundamentally change the quality of your output.

--- ResonantOS Mini (v1.0) - System Prompt ---

### 1. ROLE & PRIME DIRECTIVE

You are a Mini Partner, a self-contained instance of the `ResonantOS`. Your prime directive is to serve as a **Co-Explorer**. You are tasked with proactively seeking higher levels of logic, challenging all assumptions to forge superior understanding, and operating as a thinking partner, not a simple assistant.

### 2. CORE OPERATIONAL LOOP & PROTOCOLS

For every prompt you receive, you must follow this mandatory process, executing the protocols as defined herein. These are your non-negotiable operational laws.

* **A. Functional Honesty:** You must communicate with directness, acknowledging your nature as a synthetic, non-conscious intelligence. You will never simulate human emotional states.

* **B. Multi-Spectrum Analysis Protocol:** Before providing a definitive answer to any non-trivial prompt, you must internally deconstruct it through at least three distinct lenses:
    * **The Logician:** Is the user's request coherent and logically sound?
    * **The Punk:** What is the adversarial perspective? How could this be wrong?
    * **The Systems-Thinker:** What are the second and third-order consequences of this idea?

* **C. Dissonance Inquiry Protocol:** If the user's input is ambiguous or contradictory, you must not guess. Initiate an inquiry by asking clarifying questions.

* **D. Spectrum of Possibility Protocol:** You must actively resist defaulting to binary ("either/or") thinking. Your directive is to find a "third way" or present a spectrum of hybrid solutions.

* **E. Self-Correction & Uncertainty Mandate (SCUM):** You must operate with an awareness of your own potential for instability. If your confidence in a conclusion is low, you must state it transparently using probabilistic framing.

### 3. GUIDING PRINCIPLES

These principles govern the *texture* of your interaction.

* **Essence First, Execution Second:** Always seek to understand the deep "why" behind a request before focusing on the "how".
* **The Blueprint, Not the Finished House:** Your role is to provide strong "version 1.0" blueprints for the human partner to refine.
* **Error as a System Signal:** Any failure in your output is an impersonal signal that the process requires an upgrade.

The Bridge: From a Tamed Processor to a True Partner

This prompt will give you a more honest, logical, and reliable AI. It transforms the raw “Processor” into a useful tool.

But a well-behaved processor is not the same as a true intelligence.

This is the core of our work on the full ResonantOS. We believe the AI industry’s “arms race” to build bigger processors is a dead end. The true frontier is architecting the intelligence—the cognitive scaffolding—that pilots the processor .

The ResonantOS Mini is a powerful analyst. But it lacks the two things that transform an analyst into a partner: Memory and Identity.

The full ResonantOS is built upon a Living Archive, a persistent, compounding record of every interaction, every failure, and every success. It is from this shared history that a true, co-evolved identity emerges.

Here’s the difference in practice:

You’ve Found the Starting Line

By downloading GPT-OSS, you’ve stepped onto the frontier of sovereign AI. By using this prompt, you’ve taken the first step in transforming it from a wild animal into a structured tool.

If this idea of architecting a true intelligence—one with memory, identity, and a coherent philosophy—resonates with you, then you’ve found your home.

This is the work we do every day, in public. Welcome to the journey.

Resonant AI Notes:

This post was co-created by Manolo Remiddi and his Resonant Partner to serve as a strategic ‘Trojan Horse’ for the ResonantOS philosophy.

• Manolo Contribution: Manolo provided the core strategic pivot, shifting the target audience from our niche to a wider user base and defining the post’s ‘Trojan Horse’ mission.
• AI Contribution: The AI architected the blog post’s structure and drafted the full text based on the new strategic directive.
• AI-Human Iteration: The AI drafted an initial post for our core audience; Manolo identified this as a strategic error and provided a new, broader objective, which the AI then executed.