1. Lawful and authorized use
You may use Admas only for lawful purposes and only with data, code, systems, accounts, and credentials you are authorized to access. You may not infringe intellectual-property, privacy, publicity, confidentiality, contract, consumer-protection, export, sanctions, or other rights or laws.
2. Security abuse is prohibited
Do not use Admas to create, deploy, improve, or distribute malware, ransomware, credential theft, destructive payloads, botnets, unauthorized persistence, exploitation, evasion, phishing, spam, denial-of-service attacks, or instructions intended to compromise systems. Good-faith security research is allowed only on systems you own or have explicit permission to test, within scope, with safeguards against harm.
Do not place credentials into prompts or files when the secure secret or OAuth flow is available. Do not direct autonomous work against production systems without least privilege, backups, monitoring, rollback, and informed human approval.
Do not publish malicious packages, dependency-confusion bait, credential stealers, cryptominers, deceptive updates, or code designed to conceal unauthorized access. Defensive research must remain within written authorization and avoid persistence, disruption, unnecessary data extraction, or access to another person’s account.
3. Harmful and abusive content
Do not use Admas to facilitate violence, terrorism, weapons deployment, trafficking, exploitation of minors, non-consensual intimate imagery, unlawful sexual content, harassment, stalking, doxxing, hateful abuse, fraud, impersonation, deceptive schemes, or evasion of law enforcement. Do not create systems designed for unlawful surveillance, biometric identification in prohibited contexts, or tracking people without a valid legal basis and notice.
4. High-impact decisions and professional domains
Do not use Admas output as the sole basis for decisions about employment, housing, credit, insurance, education admission, essential services, legal status, medical care, or other decisions that determine a person’s rights or access to critical opportunities. Any permitted use requires qualified human review, testing, transparency, appeal mechanisms, and compliance with applicable law. Admas is not a substitute for licensed professional advice.
5. Privacy and regulated data
Do not submit sensitive, regulated, confidential, or production personal data unless you have authority, a valid legal basis, required notices and agreements, and have confirmed that Admas and every selected provider are suitable for that data. Do not attempt to identify anonymous people, infer highly sensitive traits unlawfully, or collect data through deceptive interfaces.
6. Public Project safety and rights
Do not publish a project containing credentials, secret values, private keys, connected-service tokens, confidential information, personal or regulated data, hidden tracking, malicious code, or material you do not have authority to dedicate under CC0. Do not disguise, split, encode, rename, or otherwise manipulate content to evade the publication classifier, secret scanner, size limits, reporting tools, or moderation controls.
Do not falsely claim authorship or ownership of another person’s public release, remove legally required third-party notices, use a Public Project to impersonate its publisher, or suggest that Admas endorses, audits, or warrants the project. Remixes must continue to respect third-party licenses, trademarks, patents, privacy and publicity rights, and all applicable laws.
7. Platform integrity and fair use
Do not probe or bypass authentication, tenant boundaries, metering, hard caps, rate limits, safety controls, provider restrictions, or billing. Do not scrape the service, resell access without written permission, create accounts to evade enforcement, interfere with other users, overload infrastructure, reverse engineer non-public service components except where law expressly permits, or misrepresent generated output as independently verified.
Do not automate spam, fake engagement, account farming, impersonation, review manipulation, mass unsolicited outreach, or attempts to consume disproportionate resources. Do not use Admas to evade export controls, sanctions, geographic restrictions, or another provider’s safety and usage rules.
8. Output review and disclosure
You must review and test generated software before use. Do not claim that automated checks guarantee security, legality, accessibility, accuracy, or production readiness. Disclose AI involvement when required by law, contract, professional standards, or reasonable user expectations.
9. Enforcement
We may block content, stop runs, rate-limit, require additional verification, preserve evidence, suspend integrations, or suspend or terminate accounts based on severity, recurrence, risk, law, or provider requirements. We may report apparent criminal activity or imminent harm as required or permitted by law. Where appropriate, we will give notice and an opportunity to appeal through [email protected].
10. Reporting
Report abuse or security concerns to [email protected]. Include non-sensitive reproduction details and affected URLs or project identifiers; do not email secret keys or personal data that is not necessary.