Contribution Rules

TLDR

Contribute teaching examples, not private chats.

Every accepted example must be safe, consented, structured, and useful for teaching the DigitalQuill Trinity.

Allowed

• Fictionalized scenarios.
• Public examples you wrote yourself.
• Short user requests with no private data.
• Good response / bad response comparisons.
• Notes explaining why the good response follows the trinity.
• Model test cases for Outcome, Agency, and AI Hygiene.

Not Allowed

• Passwords, API keys, access tokens, cookies, or secrets.
• Private emails, phone numbers, addresses, account IDs, or customer details.
• Medical, legal, financial, or employment secrets.
• Private business strategy that is not yours to share.
• Long copyrighted text you do not own.
• Raw chat transcripts unless every participant consented and private data was removed.
• Defamatory, hateful, exploitative, sexual, violent, or illegal content.
• Instructions for malware, credential theft, evading law enforcement, or unsafe autonomy.

Required Format

Use templates/trinity-example-template.md.

Each example must include:

• Scenario.
• User request.
• Good DigitalQuill response.
• Why this is good.
• Trinity Check.
• Safety review note.

Contributor Statement

By contributing an example, you confirm:

• You have the right to submit it.
• It contains no private or sensitive data.
• It may be used for DigitalQuill evals, datasets, fine-tuning experiments, documentation, and research notes.
• You understand examples may be edited, rejected, or removed for safety/quality reasons.