Terms of Service

About these terms#

These terms are an agreement between you and Allavanco Comercial Ltda., CNPJ 42.153.299/0001-64 ("we"), for the use of AI Context Orchestrator ("the platform"). If you use the platform for an organization, you accept them on that organization's behalf and confirm you're allowed to. The Privacy Policy explains how we handle personal data.

⟦How a person accepts these terms (for example a box at first sign-in) and which language version prevails⟧.

The service#

The platform lets an organization design multi-agent AI workflows on a visual Canvas, connect the services those agents may use, run the workflows through Anthropic's Claude API with the organization's own key, require a person's approval before chosen steps, and export the workflows as files for Claude Code. It also offers a Co-Pilot that proposes workflows, public apps, sharing inside the organization and a program (the Bridge) that lets a step read a folder on a paired computer.

The platform is in a controlled beta: access is by invitation, some features are incomplete or switched off, and the known limits page lists what isn't ready yet.

Eligibility and accounts#

• You must be ⟦minimum age⟧ or older and able to enter into a contract.
• You sign in with a Google or Microsoft account whose e-mail address the provider has confirmed. During the beta, you need to be on the beta list or invited by an organization's admin.
• Keep your sign-in account secure. You're responsible for what happens under your account, and you must tell us at contato@allavanco.com.br if you think someone else used it.
• Give accurate information, and don't sign in as someone else.

Organizations, roles and administrators#

Work happens inside an organization. Each member has a role that decides what they can do. Admins invite and remove members, change roles, connect services, decide what reaches Production and what's shared outside the organization.

The organization is responsible for its admins' decisions, for who it invites and for the services and keys it connects. An organization keeps at least one active admin.

Acceptable use#

You must not use the platform to:

• break the law or the rights of others, including privacy, data protection and intellectual property;
• process personal data without a legal basis, or connect a service or data you aren't authorized to use;
• build workflows that deceive people, impersonate someone, send spam, or produce harmful or illegal content;
• break the usage policies of Anthropic or of the services you connect;
• try to reach another organization's data, get around permissions, limits or security controls, or probe, scan or overload the platform;
• reverse-engineer the platform except where the law allows it, or resell access without our written permission;
• upload malicious code, or use a paired computer to read folders you aren't entitled to.

We may remove content, switch off a public app or suspend access that breaks these rules, as described below.

Your content#

Your content is what you or your organization put into the platform: workflows, prompts, instructions, References, configurations, documents, and what runs produce. It stays yours (or your organization's). You give us only the permission needed to store, process and transmit it to provide the platform to you, including sending it to Anthropic and to the services you connect when you run a workflow.

You're responsible for your content and for having the rights and legal bases to use it, including any personal data it contains. When you share a package or publish a public app, you confirm you may share it, and other people may use it as the sharing feature describes.

AI-generated output and your decisions#

Agents' answers are produced by an AI model and can be wrong, incomplete, out of date or inappropriate. Review output before you rely on it, especially for legal, medical, financial, employment or safety decisions.

You decide what agents may do. Steps that change data in a connected service run only with the tools your organization switched on, and an admin can require a person's approval before them. You're responsible for the actions you authorize and for their consequences in your connected services.

AI providers, integrations and third-party services#

Live runs use your organization's own Claude API key: your use of Claude is also governed by your agreement with Anthropic, and Anthropic bills the tokens to that key. The services you connect (for example Google, Slack, GitHub) are governed by their own terms. We don't control those services and aren't responsible for their availability, their changes or how they handle data you send them.

Public apps and sharing#

An admin can publish a workflow as a public app that anyone with the link can use; visitors' answers run on the organization's key and count against its limits. An admin can share packages inside the organization and submit them to the Community, where our reviewers may approve, reject or take them down. Using someone else's package never runs anything until you choose to, and you're responsible for checking it before you do. Some of these features are switched on per server and may not be available to you.

Paired computers#

If you pair a computer through the Bridge, you're responsible for that computer, for the folders you allow, and for the Claude Code installation and account it uses. A step only reads the folders the computer's policy names.

Plans, billing, cancellation and refunds#

During the beta, organizations use the beta plan at no charge, and tokens are billed by Anthropic to the organization's own key. When paid plans are offered, prices and limits will be shown before you subscribe, payments will be processed by Stripe, and subscriptions will renew until cancelled.

⟦Cancellation, refund and price-change terms for paid plans, including the right of withdrawal under the Consumer Defense Code where it applies⟧.

Availability, beta limitations and changes#

The platform is offered as a beta: it may change, have errors, be unavailable at times, or lose a run in progress (runs are held in memory, so a restart ends them). We may add, change or remove features, and we'll tell you ahead of changes that take away something you rely on, where we reasonably can. ⟦Any service level the company commits to⟧.

Keep your own copies of what matters: you can export your workflows at any time.

Suspension and termination#

You can stop using the platform at any time; an admin can ask us to close the organization. We may suspend or end access if these terms are broken, if required by law, to protect users or the platform, or when the beta ends, ⟦notice period and how the company will notify⟧.

Closing an account and what happens to data#

When an account or an organization is closed, we delete its data as described in the Privacy Policy: our team does it on request for now. Data can remain in database copies until they expire (up to 60 days), and records the law requires us to keep are kept for the required period. Before closing, an admin can ask for a copy of the organization's data. ⟦How long after closing a copy can still be requested⟧.

Intellectual property#

The platform, its software, design, templates we provide and the name AI Context Orchestrator belong to us or our licensors. These terms give you the right to use the platform while they're in force; they don't transfer ownership.

Confidentiality#

We treat your organization's content and credentials as confidential and use them only to provide the platform, as the Privacy Policy describes. ⟦Any further confidentiality commitments, in both directions⟧.

Disclaimers#

⟦Disclaimer of warranties, as far as Brazilian law allows⟧.

Limitation of liability#

⟦Limitation of liability, as far as Brazilian law allows⟧.

Governing law and disputes#

⟦Governing law, competent court (foro) and any step to try to settle a dispute first⟧.

Changes to these terms#

We may update these terms. We'll show the new version here with its date and tell you in the platform before a material change applies. ⟦How a person accepts a new version, and what happens if they don't⟧.

Contact#

Allavanco Comercial Ltda. · CNPJ 42.153.299/0001-64 · contato@allavanco.com.br · Travessa Maria Quitéria, 110, São José dos Campos, São Paulo, Brazil.

Version history#