Effective date: October 20, 2026
CONTENT
1.
Agreement and acceptance
2.
Definitions
3.
Accounts and Hubspaces
4.
The software
5.
Your content
6.
Outputs
7.
Cogs and Actions
8.
Execution on your device
9.
Connectors and Third-Party Services
10.
Inference Sources
11.
Acceptable use
12.
Privacy
13.
Term and termination
14.
Disclaimers
15.
Limitation of liability
16.
Indemnification
17.
Changes to these Terms
18.
Limitation on time to file claims
19.
Governing law and dispute resolution
20.
General
These Terms of Service (the “Terms”) are entered into between you and OpenTeams, Inc. (“OpenTeams,” “we,” “us”) and govern your access to and use of the Collab desktop application, the hosted Hub service, and the related software, services, and documentation we make available (collectively, the “Services”).
By downloading, installing, or using the Services, or by clicking to accept these Terms, you agree to be bound by them and by our Privacy Policy at https://openteams.com/collab-privacy-policy and our Acceptable Use Policy at https://openteams.com/collab-acceptable-use-policy, each incorporated by reference. If you do not agree, you may not use the Services.
These Terms take effect on the Effective Date stated above.
If you are entering into these Terms on behalf of an entity, you represent and warrant that you have authority to bind that entity, and you agree on that entity’s behalf to be bound, with “you” and “your” referring to you, that entity, and other users accessing the Services on that entity’s behalf.
You must be at least 18 years old, or the age of majority in your jurisdiction if higher, to use the Services.
“Action” means any operation a Cog performs at your direction or on your behalf — whether you direct it in a conversation, invoke it, schedule it, or authorize it as part of a larger piece of work — including reading or retrieving data through a Connector, executing a command in a Sandbox, reading or writing files in a Workspace, or calling a tool.
“Cog” means a configured AI worker made available through the Services, comprising a model, context (including Frames), tools, and permissions.
“Connector” means an authenticated integration between the Services and a Third-Party Service you link to your account.
“Connector Content” means data the Services retrieve from a Third-Party Service through a Connector.
“Frame” means a scoped, text-based context artifact you create, store, or share through the Services.
“Collab Hub” means the hosted service through which accounts, Hubspaces, Frames, groups, and shared resources are stored and administered.
“Hubspace” means an organizational tenancy within the Hub.
“Inference Source” means a model provider used to generate Outputs, which may be (a) hosted by us, (b) running locally on your device, or (c) a third-party provider you connect using your own credentials.
“Input” means content you submit to the Services, including prompts, files, Frames, and Skills.
“Output” means content the Services generate in response to an Input.
“Sandbox” means an isolated execution environment, which may run on your device using a container runtime you supply, in which Cogs execute commands and tools.
“Skill” means a named, installable unit of instructions and accompanying scripts that a Cog may load and execute through the Sandbox.
“Third-Party Service” means any service, software, model provider, or content source not provided by us, including services reached through a Connector or an Inference Source you configure.
“Workspace” means a directory on your device or in a Hubspace that the Services make available to a Cog.
You are responsible for safeguarding your credentials and for all activity under your account. Notify us promptly of any unauthorized use.
A Hubspace may be provisioned by an organization — typically your employer, or another entity that has agreed with us to make the Services available to its members (a “Hubspace Administrator”). A Hubspace Administrator is a third party, separate from OpenTeams.
If you use the Services through a Hubspace you did not provision yourself, the Hubspace Administrator, not OpenTeams, controls that tenancy. It may:
We act on a Hubspace Administrator’s instructions with respect to its tenancy. Any dispute about a Hubspace Administrator’s decisions — including a decision to access or delete your content, or to end your access — is between you and that organization, and we are not responsible for those decisions.
This section governs the Hubspace Administrator’s access. OpenTeams’ own access to content is separate and is governed by §12 (Privacy), our Privacy Policy, and any applicable Data Processing Addendum.
We may suspend or terminate access, in whole or in part, for violation of these Terms, for security reasons, or where required by law.
Subject to these Terms, we grant you a limited, revocable, non-exclusive, non-transferable, non-sublicensable license to install and use the Collab desktop application on devices you own or control, for your internal business or personal use.
You may not (a) reverse engineer, decompile, or disassemble the Services except to the extent that restriction is unenforceable under applicable law; (b) rent, lease, sell, sublicense, or provide the Services as a service bureau; (c) remove or obscure proprietary notices; (d) circumvent or disable security, authorization, sandboxing, or rate-limiting features; or (e) use the Services to develop a competing product or to train a competing machine learning model.
The Services include an automatic update mechanism. You consent to our downloading and installing updates, including security updates, without further notice. We may require a minimum version and may disable or degrade materially outdated versions.
The Services incorporate and distribute third-party and open-source components, including bundled language and package runtimes, an agent runtime, and container images obtained from public registries. Those components are licensed to you under their own license terms, which are provided in the Collab documentation https://collab.openteams.dev/hub/ and which control over these Terms with respect to those components. Nothing in these Terms grants you rights in third-party components beyond what their licenses grant, and we make no warranty with respect to them.
Machine learning models you obtain and use with the Services are licensed under their own terms, which may impose restrictions on use, including field-of-use limits, acceptable-use riders, and deployment thresholds. You are responsible for compliance with the license of each model you use.
Certain features require software we do not provide. A Cog that runs in a Sandbox requires a container runtime; a Cog configured for direct host execution does not. You are responsible for obtaining, licensing, and maintaining any such software. Some such software is separately licensed by its vendor and may require a paid subscription for your organization; you are responsible for that compliance.
As between you and us, you retain all right, title, and interest in your Inputs, Frames, Skills, and Workspace contents. We claim no ownership in them.
You grant us a worldwide, non-exclusive, royalty-free license to host, store, reproduce, transmit, display, and create derivative works of your Inputs solely to (a) operate, provide, secure, and support the Services for you; (b) where you direct sharing, make the content available to the recipients you designate; and (c) comply with law. This license terminates when the content is deleted, except for (i) content you have shared, as to the recipients who received it, and (ii) backups retained for a commercially reasonable period.
We will not use your Inputs, Outputs, Frames, Skills, Connector Content, or Workspace contents to train or improve generalized machine learning or artificial intelligence models, except where you have given separate, affirmative, revocable consent. We may use aggregated or de-identified operational data that does not identify you or your content to maintain and improve the Services.
If you send us suggestions or feedback, we may use them without restriction or obligation to you.
Subject to your compliance with these Terms, we assign to you all of our right, title, and interest, if any, in Outputs generated for you.
Outputs are not unique. Other users may receive identical or similar Outputs, and our assignment in §6.1 extends only to the Outputs generated for you.
Outputs may be inaccurate, incomplete, out of date, or offensive, and may misstate facts about people, places, code, or events. Outputs are not professional advice. You must independently verify any Output before relying on it, and you are solely responsible for any decision you make or action you take based on an Output. Outputs may reflect biases present in the data a model was trained on. Where the law requires you to disclose that content you publish or share was generated by AI, that disclosure is your responsibility.
The Services are general-purpose tools for knowledge work and software development, intended to be used under human direction.
Cogs take Actions at your direction and on your behalf. You bear full responsibility for all Actions taken on your behalf, including Actions that read, modify, transmit, or delete data, execute code, or interact with Third-Party Services. You represent that each Action you authorize complies with these Terms, with applicable law, and with any agreement governing the systems and data the Action touches.
You determine which Cogs are available to you, and — through a Cog’s own definition and the settings you apply to it — the instructions it works under, the context and Frames it applies, the model it uses, the Skills, tools, and Connectors it may call, the Workspaces it may access, and the limits placed on it. You are responsible for configuring these appropriately for the sensitivity of the data involved and the consequences of the Actions the Cog may take. Where you adopt a Cog defined by someone else, you remain responsible for what it does on your behalf.
A Cog may take Actions you did not intend, misinterpret its instructions, repeat or fail to complete an Action, or produce work that is incorrect or incomplete — including where it runs without a person watching. Do not authorize a Cog to take an Action whose consequences you are not prepared to bear, and maintain independent backups of data a Cog can modify or delete.
Content retrieved from Third-Party Services, Workspaces, or other external sources may contain instructions intended to manipulate a Cog into taking Actions you did not authorize or into disclosing information. The Services apply safeguards against this, but no safeguard is complete. We do not warrant that a Cog will disregard adversarial instructions embedded in external content, and you assume the risk of Actions influenced by such content.
Certain features execute code on your device. Each Cog is configured, before it runs, to execute its tools either (a) in a Sandbox — a container created through a container runtime you supply, with a directory from your device mounted into it — or (b) directly on your device, outside any Sandbox. The Sandbox is the default. The Execution Mode is fixed when the Cog is configured and cannot be changed afterward. Skills you install run under whichever mode the Cog uses.
You represent and warrant that, on each device where you do so, you are authorized to install the Services, to run a container runtime where a Sandbox is used, to execute programs directly on the device where direct host execution is used, and to grant access to the directories you designate as Workspaces. If the device is managed by an employer or other organization, you are responsible for confirming that this use is permitted by that organization’s policies — direct host execution in particular may not be.
Sandboxing reduces but does not eliminate risk. We do not warrant that a Sandbox will prevent code from accessing, modifying, or transmitting data outside it, from consuming device resources, or from reaching your local network. Isolation depends on software we do not control, including your container runtime and operating system. You assume the risk of executing code in a Sandbox, and you should not designate as a Workspace any directory whose contents you are not prepared to expose to executed code.
Where you configure a Cog to execute directly on your device, there is no Sandbox and no isolation boundary. Its tools run as processes on your operating system with your user account’s privileges, and can read, modify, create, or delete anything that account can reach — inside and outside the Workspace — and can use your network connection.
The Services provide controls for this mode, and configuring them is your responsibility: a setting that can require your approval before a tool runs, and restrictions on writing into certain protected directories. These controls are partial. The write restrictions apply to particular file-writing tools and do not constrain a tool that executes shell commands, which can write anywhere your account can write. If you set the approval setting to allow tools to run without approval, they will run unattended.
The Services provide controls for this mode, and configuring them is your responsibility: a setting that can require your approval before a tool runs, and restrictions on writing into certain protected directories. These controls are partial. The write restrictions apply to particular file-writing tools and do not constrain a tool that executes shell commands, which can write anywhere your account can write. If you set the approval setting to allow tools to run without approval, they will run unattended.
Select this Execution Mode only where you accept that outcome, and do not select it on a device holding data you are not prepared to expose to code the Cog runs.
Execution on your device and local Inference Sources may consume substantial CPU, memory, disk, and power. You are responsible for any resulting cost or device impact.
When you link a Connector, you authorize the Third-Party Service to issue a credential that is held by our hosted Hub service and used to make requests to that service on your behalf. You are not granting access to the Collab desktop application or to any Cog: the desktop application, Cogs, Skills, and code running in a Sandbox never receive that credential. They receive only the data returned in response to a request.
Our hosted Hub service uses the credential to call the Third-Party Service only in response to a request originating from your session — when you, or a Cog acting at your direction, invoke a feature that requires it — and retrieves only the data responsive to that request. We do not use the credential to access the Third-Party Service on a standing, scheduled, or background basis, except where a feature you expressly enable does so and says so.
The authorization persists until you disconnect the Connector in the Services or revoke it at the Third-Party Service. Disconnecting stops further access. It does not delete Connector Content already retrieved, which may remain in your conversation history, Workspaces, or other local or Hub storage until you delete it.
The permissions granted at connection may be broader than any single feature requires, and a single authorization may cover more than one Connector. The Services describe, at the point of connection, what is being requested. You are responsible for reviewing that disclosure.
Connector Content retrieved at your direction becomes part of your Inputs and is handled under §5. In addition, and notwithstanding anything else in these Terms: we use Connector Content only to provide the user-facing features you invoke; we do not sell it; we do not use it to train or improve generalized machine learning or artificial intelligence models; and we do not permit human review of it except with your affirmative consent, for security or abuse investigation, or where required by law. Where a Third-Party Service’s own developer policy imposes stricter limits, those limits also apply.
Your use of a Third-Party Service is governed by your agreement with its provider. We do not control Third-Party Services and are not responsible for their availability, accuracy, security, or data practices, or for any loss arising from your use of them. Your connection may be subject to the Third-Party Service’s terms, and a provider’s change may break or disable a Connector without notice.
You choose which Inference Source processes your Inputs. Your choice determines where your content is sent.
Where we provide inference, we transmit your Inputs to our model-hosting providers you authorize under contracts that restrict their use of that content. Our Privacy Policy identifies these providers.
Where you configure a model running on your own device, Inputs sent to it do not leave your device through the Services.
You may configure a third-party provider using your own credentials. If you do, you contract directly with that provider; you are responsible for your agreement with them, for all charges they impose, and for their handling, retention, and training use of the content you send. We transmit your Inputs — which may include Connector Content — to that provider at your direction, and we are not responsible for what they do with it. Review that provider’s terms before enabling this option, particularly if you have enabled Connectors.
Inference Sources may change. We may add, remove, or substitute models and providers, and a model available today may not be available later.
Your use of the Services is subject to our Acceptable Use Policy (Schedule A). We may investigate suspected violations and may remove content, suspend access, or terminate accounts in response. We may also cooperate with law enforcement as required by law.
Our collection and use of personal information is described in our Privacy Policy at https://openteams.com/collab-privacy-policy/
We may suspend or terminate your access (i) immediately at any time for breach of these Terms, or for security or legal reasons, or (ii) on 30 days’ notice for convenience. We may terminate accounts inactive for more than one year on notice.
On termination, your license under §4.1 ends and you must cease using and remove the software. Content stored in the Hub may be deleted after 90 days. Content stored locally on your device — including conversation history, Workspace files, installed Skills, and provisioned environments — is not removed by termination and remains your responsibility to manage.
Sections 2, 5.1, 5.4, 6, 7.1, 8.3, 9.6, 10.4, and 14–20 survive termination.
THE SERVICES, OUTPUTS, AND ACTIONS ARE PROVIDED ON AN “AS IS” AND “AS AVAILABLE” BASIS, WITHOUT WARRANTIES OF ANY KIND, EXPRESS, IMPLIED, STATUTORY, OR OTHERWISE. TO THE FULLEST EXTENT PERMITTED BY LAW, WE DISCLAIM ALL WARRANTIES, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, NON-INFRINGEMENT, AND ANY WARRANTY ARISING FROM COURSE OF DEALING OR USAGE OF TRADE.
WITHOUT LIMITING THE FOREGOING, WE DO NOT WARRANT THAT THE SERVICES WILL BE UNINTERRUPTED, SECURE, OR ERROR-FREE; THAT OUTPUTS WILL BE ACCURATE OR SUITABLE FOR ANY PURPOSE; THAT ACTIONS WILL BE PERFORMED CORRECTLY OR COMPLETELY; THAT A SANDBOX WILL CONTAIN EXECUTED CODE, OR THAT CODE RUN OUTSIDE A SANDBOX WILL BE CONSTRAINED AT ALL; OR THAT CONNECTORS WILL REMAIN AVAILABLE.
THE FOREGOING DOES NOT AFFECT WARRANTIES THAT CANNOT BE EXCLUDED UNDER APPLICABLE LAW.
TO THE FULLEST EXTENT PERMITTED BY LAW, NEITHER PARTY WILL BE LIABLE FOR INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, EXEMPLARY, OR PUNITIVE DAMAGES, OR FOR LOST PROFITS, LOST REVENUE, LOST GOODWILL, OR LOSS, CORRUPTION, OR DISCLOSURE OF DATA, EVEN IF ADVISED OF THE POSSIBILITY.
OUR TOTAL AGGREGATE LIABILITY ARISING OUT OF OR RELATING TO THESE TERMS OR THE SERVICES WILL NOT EXCEED THE GREATER OF THE AMOUNTS YOU PAID US IN THE SIX (6) MONTHS PRECEDING THE EVENT GIVING RISE TO THE CLAIM OR USD $100.
THESE LIMITATIONS DO NOT APPLY TO LIABILITY ARISING FROM GROSS NEGLIGENCE OR WILLFUL MISCONDUCT, OR TO LIABILITY THAT CANNOT BE LIMITED UNDER APPLICABLE LAW.
You will defend, indemnify, and hold harmless OpenTeams, its affiliates, licensors, and service providers, and their respective officers, directors, employees, and agents, from and against any claims, liabilities, damages, losses, and expenses (including reasonable attorneys’ fees) arising out of or relating to: (a) your breach of these Terms or the Acceptable Use Policy; (b) your Inputs, Frames, Skills, or Workspace contents; (c) Actions taken on your behalf; (d) your use of a Third-Party Service or an Inference Source you configure; (e) your violation of law or of the rights of a third party; or (f) your failure to hold the authority represented in §8.2
We may revise these Terms. For material changes, we will provide at least 30 days’ notice by in-app notice or email before they take effect. Changes apply prospectively. Your continued use after the effective date constitutes acceptance. If you do not agree, you must stop using the Services.
ANY CLAIM ARISING OUT OF OR RELATING TO THESE TERMS OR THE SERVICES MUST BE COMMENCED WITHIN ONE (1) YEAR AFTER THE CAUSE OF ACTION ACCRUES, OR IT IS PERMANENTLY BARRED, EXCEPT WHERE A LONGER PERIOD IS REQUIRED BY APPLICABLE LAW.
These Terms are governed by the laws of the State of Texas, without regard to conflict-of-laws rules.
Subject to §20.3, any suit must be brought exclusively in the federal or state courts located in Austin, Travis County, Texas, and the parties consent to personal jurisdiction there.
If any provision is held unenforceable, it will be limited or severed to the minimum extent necessary and the remainder will continue in effect. No waiver is effective unless in writing, and no failure to enforce is a waiver.
You may not assign these Terms without our prior written consent. We may assign them in connection with a merger, acquisition, or sale of assets.
You represent that you are not located in, and will not use the Services in, a jurisdiction subject to comprehensive sanctions, and that you are not a person with whom transactions are prohibited under applicable export-control or sanctions laws.
Notices to us shall be sent to: [email protected]. Notices to you may be given in-app or to the email associated with your account. You consent to receive communications, notices, disclosures, agreements, and other documents relating to these Terms (collectively, “Communications”) in electronic form, including by email to the address associated with your account. You agree that all communications provided electronically satisfy any legal requirement that such communications be in writing, to the same extent as if provided in paper form. A communication sent by email shall be deemed received on the date sent, if sent before 5:00 p.m. Central Time, and otherwise on the next business day. You agree to promptly notify us in writing of any change to your designated email address.
Neither party is liable for delay or failure to perform due to causes beyond its reasonable control.
You can report inaccurate, harmful, or unexpected Outputs or Actions, security vulnerabilities, and suspected misuse of the Services to [email protected]. We review reports to identify and correct problems with the Services. If an incident materially affects the security of your content, we will notify you as described in our Privacy Policy.
