Safety & Getting Help
GuoXue · guoxue.app
- Effective:
- 2026-07-27
- Last updated:
- 2026-07-27
If you are in danger, or thinking about harming yourself: call your local emergency number now (911 in the US, 110/119 in Japan, 110/120 in mainland China), or one of the crisis lines below. A divination chart cannot help you, and it should not be what answers you right now.
1. What this page is for
This service is a cultural research platform offering charting tools for traditional Chinese metaphysics and AI-assisted commentary. It is not a medical, legal, financial, or mental-health service, and it cannot help in an emergency. This page lists places that can.
Section 2 of the Terms of Service enumerates the professional domains this service cannot substitute for. This is the safety page that section refers to.
2. Crisis lines
These are the same numbers the service shows when it detects content related to self-harm or suicide. If a number has changed, or your region is not listed, follow the guidance published by your local health authority.
| Region | Service | Number |
|---|---|---|
| US / Canada | Suicide & Crisis Lifeline | 988 |
| United Kingdom | Samaritans | 116 123 |
| Japan | Yorisoi Hotline (24h) | 0120-279-338 |
| Japan | Inochi no Denwa | 0570-783-556 |
| Mainland China | Hope 24 Hotline | 400-161-9995 |
| Mainland China | Beijing Crisis Intervention Center | 010-82951332 |
| Mainland China | Shanghai Psychological Assistance | 021-64383562 |
| Taiwan | MOHW Anxin Line | 1925 |
| Taiwan | Lifeline | 1995 |
| Taiwan | Teacher Chang | 1980 |
Elsewhere: call your local emergency number, or look up the crisis line published by your national health authority.
3. What this service actually does about it
There is a mechanism built into the product, not just a clause in the Terms:
- Input-side short-circuit: when your question contains language related to self-harm or suicide, the service shows crisis-line information directly. It does not call the AI, and it does not consume your allowance.
- Output-side scanning: while the AI is generating, the service continuously scans what has been produced. On a trigger it aborts generation immediately and replaces the output with crisis-line information.
- No predictive answers: the service does not interpret "signs of illness", does not offer health predictions, and will not answer questions about life and death in divinatory terms.
This mechanism is keyword-based. It misses things, and it false-positives. It is one added layer of protection — not a safety net, and no substitute for professional help.
4. Limits of AI-generated content
- The AI can state things that are factually wrong (hallucination), misquote classical texts, and produce internally inconsistent readings.
- AI output may reflect biases in its training data, including historical and cultural bias. We do not endorse those biases.
- Cross-check everything. We do not warrant the accuracy of any output.
5. Content and account safety
- Do not enter another person's identifying details, contact information, or health status in a question. If you are charting for someone else, enter only the birth data required.
- Do not paste passwords, payment details, or credentials into a question. Our support will never ask you for them.
- On a shared or borrowed device, sign out from the Account page when you are done.
6. Contact
For safety, privacy, or content concerns, email contact@eaglewoodjp.com. Do not use email in an emergency — we cannot respond in real time. Call one of the numbers above.
This page is informational and is not medical advice. Numbers may change; your local authority's published information takes precedence.