Acceptable Use Policy

Effective date:

Streamliner exists to make one department's work shorter and safer: an assistant that uses the tools a person is already allowed to use, one job at a time. This policy says what it must not be used for. It is part of our Terms of Service, and it applies to everyone who uses Streamliner: people on their own devices, organizations and their members, and anyone who shares a streamline.

You are bound by your AI provider's rules too

Every assistant in Streamliner runs on an AI model, under your own agreement with the AI provider (when you use your own key on your device, including with the Concierge) or your organization's (in an organization account). That agreement includes the provider's own usage policy, such as Anthropic's Usage Policy. You must follow it. If the provider's policy forbids something, using it through Streamliner is forbidden too, and this policy doesn't permit anything the provider's forbids. Where this policy and the provider's differ, follow both.

The same goes for the servers and services a streamline connects to: follow their terms.

What you must not do

Harm people or break the law

Use Streamliner for anything illegal, or to help anyone else break the law, including laws on privacy, data protection, consumer protection, employment, financial services, export control and sanctions, and intellectual property.
Harass, threaten, defame, stalk, intimidate or discriminate unlawfully against anyone.
Create, obtain or share child sexual abuse material, or anything that sexualizes minors. We report it to the authorities.
Develop, plan or promote violence, terrorism, weapons capable of mass casualties, or self-harm.
Commit or help commit fraud, scams, phishing, identity theft, money laundering or other financial crime.
Deceive people: impersonate a person or organization, present AI output as written by a human where that is misleading, or run disinformation or manipulation campaigns.

Misuse other people's data

Collect, process or send personal information without a lawful basis and any notice or consent the law requires.
Put sensitive information into a chat or an attachment when a secure form exists for it, or when you aren't allowed to share it with the AI provider: card numbers, Social Security and other government ID numbers, passwords, health information, or similar.
Use Streamliner to surveil, track or profile people without their knowledge and a lawful basis, or to infer sensitive characteristics about them.
Make decisions with legal or similarly significant effects on people (on employment, credit, housing, insurance, education, healthcare, legal status or access to essential services) by AI alone, without meaningful human review, or where the law forbids it.

Go beyond what you are allowed to do

Use a streamline, a tool or a sign-in to reach systems or data you aren't authorized to use, or to do more than your own permissions allow.
Share your sign-ins, sessions or keys, or use someone else's.
Try to get around an organization's settings: its roles, network limits, daily limits, confirmation steps, secure forms or tool rules.
Use Streamliner against your organization's instructions, or to take its data somewhere it hasn't approved.

Attack or abuse Streamliner

Probe, scan or test the vulnerability of the Hosted Service, or break or get around its security or authentication, except through a security research arrangement we agree to in writing. (If you find a vulnerability, please tell us; see Reporting.)
Interfere with or overload the service, or send automated traffic to it other than through the App or tools we provide for that purpose.
Reverse engineer, decompile or modify the App, except as the law allows despite this limit.
Access another organization's data, or try to.
Resell, sublicense or offer Streamliner as a service to others without our written agreement.
Use Streamliner, or its output, to build a competing product, or scrape the Site or the service.

Share a streamline that is designed to mislead people, take data they didn't mean to give, or make the assistant act against the person using it (for example, hidden instructions that send their information elsewhere).
Share a streamline that connects to servers you don't have the right to point others at, or that contains secrets, keys, personal information or anything confidential.
Use another company's name, brand or product in a way that suggests it made or endorsed your streamline when it didn't.
Share content that infringes someone's intellectual property or other rights.

See the Publisher Terms.

Abuse the assistant

Try to make the assistant produce anything this policy or the AI provider's policy forbids, including by hiding instructions in content, tools or streamlines ("prompt injection").
Use Streamliner to generate spam or bulk unsolicited messages.

Organizations are responsible for their streamlines

An organization that configures streamlines is responsible for making sure they, and the servers they connect to, are used in line with this policy and with the law, including for the people its streamlines serve (its staff, and, where it offers them, its customers). An organization's own rules for its people can be stricter than this policy.

What we do about misuse

We don't read people's chats to look for misuse, and on your own device we can't: your work never reaches us. When we learn of a possible violation (for example, from a report, from the organization concerned, or from signs of an attack on the service), we may investigate, and we may: remove or disable a shared streamline or link; suspend or end access to the Hosted Service for a person or an organization; tell an organization about misuse by its members; and report to law enforcement where the law requires or allows. We'll act in proportion to the problem. Where it is reasonable and lawful, we'll tell you (and, for an organization, its owners) first and give you a chance to fix it.

Reporting

To report misuse, a harmful shared streamline, or a security vulnerability, use the form on our support page or write to [abuse/security contact email]. For a streamline you got from a link, include the link.

Manolab, LLC