# Terms of Use

**Effective date:** September 5, 2026  
**Applies to:** LLM Ambiguity Lab v1 and its accompanying public documentation and machine-readable specification files.  
**Project provider and responsible maintainer:** Deshimaru Sakaguchi

## 1. Free access

LLM Ambiguity Lab v1 is provided as a free research and educational resource. No payment or account is required to use the public v1 interface.

Free access does not place the project, documentation, source code, or project name in the public domain.

## 2. Permitted use

You may use the public Lab for lawful purposes including:

- personal learning;
- teaching and demonstration;
- research;
- software and agent-design discussion;
- internal product-design exploration;
- citation and commentary;
- linking to the canonical public page.

You may quote brief portions of the documentation with appropriate attribution, subject to applicable law.

## 3. Source code and redistribution

Unless a separate license file or explicit license notice states otherwise, no additional open-source license is granted merely because the browser-delivered HTML, CSS, JavaScript, Markdown, or JSON files can be viewed or retrieved.

Do not redistribute a modified version in a way that falsely represents it as the official LLM Ambiguity Lab release.

You may describe, critique, implement, or independently adapt the architectural ideas, subject to applicable intellectual-property law and any third-party rights.

## 4. Attribution

When publishing academic, technical, or substantial derivative discussion based directly on this release, attribution is appreciated and should identify **Deshimaru Sakaguchi**, **LLM Ambiguity Lab v1**, and its canonical page.

See `CITATION.md` for suggested formats.

## 5. No model-confidence or safety guarantee

The Lab is an illustrative deterministic simulation.

Its `context_resolution_score` is not an LLM's internal confidence, a calibrated probability, or a guarantee that an interpretation is correct.

The Lab does not guarantee:

- factual accuracy;
- hallucination prevention;
- safe agent execution;
- production suitability;
- recovery of a user's true internal intention.

## 6. No professional advice

The Lab and documentation are provided for research, educational, and design-exploration purposes. They are not legal, medical, financial, security, psychological, or other professional advice.

## 7. User responsibility

You are responsible for how you use concepts, examples, code patterns, or outputs from the Lab in any real system.

Before applying similar control logic in a production or high-impact environment, you should perform appropriate testing, risk assessment, authorization design, domain review, and human oversight.

## 8. Availability and changes

The project may be corrected, revised, expanded, moved, or discontinued. Features and documentation may change between versions.

When exact reproducibility matters, cite the version number and retain the relevant release files.

## 9. Third-party links

The Lab may link to related external sites, including SANA OS. External sites may have their own terms, privacy practices, and availability. A link does not make those external services part of the v1 simulator.

## 10. Privacy and local processing

As implemented in v1.0, the ambiguity simulation itself runs in the browser and does not require an LLM API call.

This statement concerns the simulator logic only. The hosting provider or website infrastructure may process ordinary request metadata or logs independently of the simulator.

Do not submit confidential or sensitive information merely to test the public demo.

## 11. Disclaimer of warranties

The Lab is provided on an **as-is** and **as-available** basis to the extent permitted by applicable law.

No warranty is made that the Lab will be uninterrupted, error-free, complete, suitable for a particular purpose, or compatible with every browser or system.

## 12. Limitation of responsibility

To the extent permitted by applicable law, the project provider is not responsible for decisions, deployments, losses, or harms resulting from treating this educational simulation as a validated production safety or intent-classification system.

Nothing in these terms excludes rights or obligations that cannot lawfully be excluded.

## 13. Version boundary

These terms apply to **LLM Ambiguity Lab v1**. A later SANA OS extension or other separate tool may have different capabilities, documentation, or terms.