You must have authority to provide each instruction, credential, system, and item of company data.
Independent model verification improves evidence; it does not replace qualified human judgment.
Use least privilege, human approval, spending limits, and provider revocation for sensitive actions.
Agreement and eligibility
These Early Access Terms of Service (“Terms”) are an agreement between Agent9 AI (“Agent9 AI,” “Rally,” “we,” or “us”) and the person or organization using Rally (“Customer” or “you”). They govern access to the Rally website, hosted administration surface, control plane, agent coordination runtime, email workflow, public evidence console, and related services (the “Service”).
By creating a connection, commissioning a run, accessing an account, or otherwise using the Service, you accept these Terms and the Privacy Policy. If you use Rally for an organization, you represent that you have authority to bind it and to administer the systems and data you connect. You must be at least 18 and legally able to enter this agreement.
If you do not accept these Terms, do not use the Service. A separately signed order, pilot agreement, enterprise agreement, data-processing agreement, or other written contract controls to the extent it expressly conflicts with these Terms.
What Rally provides
Rally coordinates separate AI models or agents around one customer-defined goal, approved company systems, deterministic authority, evidence, recovery, and independent review. Features may include Google-authenticated administration, encrypted credential custody, agent-to-agent interoperability, tool gateways, email commissioning, public run evidence, and self-hosted components.
Rally is an orchestration and governance service—not an employer, staffing agency, professional adviser, records system of final authority, or guarantee of a correct outcome. “Verified” means that Rally’s configured checks and independent-owner rule completed; it does not mean a fact, recommendation, compliance conclusion, or artifact is universally correct.
Accounts, administrators, and authority
- You must provide accurate account information, maintain authorized administrators, and protect Google, provider, email, and device credentials.
- You are responsible for activity initiated through your accounts, commissioner addresses, connected services, agents, and delegated users, except to the extent caused by Rally’s breach of these Terms.
- You may connect only accounts, repositories, workspaces, datasets, domains, payment resources, and other systems you are authorized to access and delegate.
- You must configure least-privilege scopes, resource boundaries, budgets, approval gates, human stops, retention, and publication settings appropriate to the work.
- Notify terry@agent9.dev promptly of suspected unauthorized use, exposed credentials, or a security incident involving Rally.
Rally may rely on instructions from a verified administrator or commissioner. Customer is responsible for its internal authorization, employment, records, confidentiality, and regulatory obligations.
Customer content, credentials, and public evidence
Your content remains yours
As between the parties, Customer retains its rights in goals, source material, messages, files, connected-system data, and other content submitted to the Service (“Customer Content”). Customer grants Rally a limited, non-exclusive right to host, copy, transmit, transform, and otherwise process Customer Content solely to provide, secure, support, and improve the Service and to comply with law.
You represent that you have all rights and permissions needed for Customer Content and Rally’s processing instructions. Do not submit regulated, export-controlled, highly sensitive, or third-party confidential data unless you have assessed the Service, configured appropriate safeguards, and entered any required written agreement.
Credentials
You may import or authorize credentials only through Rally’s designated connection path. Do not place credentials in prompts, emails, tasks, repositories, screenshots, or public evidence. A stored credential remains subject to its provider’s permissions and must be revoked at the provider when compromised or no longer needed. Rally may refuse overly broad, unsupported, or unverifiable credentials.
Publication
Runs are private by default. If an authorized operator deliberately enables publication and selects a run, Customer authorizes Rally to publish the resulting sanitized evidence until it is withdrawn. Customer must ensure the title, evidence, artifacts, and metadata selected for publication contain no confidential, personal, or restricted information. Public copies made by third parties may remain after Rally removes its own projection.
AI output and third-party services
AI systems can be inaccurate, incomplete, biased, insecure, unavailable, or confidently wrong. Agent handoffs, evidence receipts, different model families, and independent verification reduce some risks but do not eliminate them. You must evaluate outputs for accuracy, suitability, intellectual-property concerns, security, and legal or business consequences before relying on or distributing them.
Rally interoperates with services supplied by Google, Anthropic, OpenAI, Cloudflare, Resend, and systems selected by Customer. Those services are independent, may change or fail, and are governed by their own contracts, acceptable-use rules, privacy terms, fees, and data controls. Customer authorizes Rally to transmit the necessary part of a task to each enabled provider. Rally is not responsible for a third party’s service, output, suspension, or change.
Acceptable use
You will not, and will not help another person to:
- use the Service unlawfully, deceptively, abusively, or in violation of another person’s rights or a provider’s terms;
- access or connect systems, accounts, data, identities, or credentials without authorization;
- introduce malware, destructive instructions, credential theft, spam, harassment, surveillance, or attempts to evade safety controls;
- circumvent permissions, human approvals, budgets, model-family separation, verification rules, rate limits, or security boundaries;
- probe, scan, or test the vulnerability of the hosted Service without written authorization, except where applicable law protects good-faith security research;
- use Rally to make prohibited automated decisions about people, create illegal content, infringe intellectual property, or facilitate physical or digital harm;
- misrepresent AI-generated work as independently verified when Rally’s evidence does not support that claim; or
- resell or provide the Service to third parties without written authorization.
Rally may investigate suspected abuse, preserve relevant evidence, narrow authority, suspend a connector or account, or contact providers or authorities when reasonably necessary to protect the Service or comply with law.
Ownership, open source, and feedback
Agent9 AI and its licensors retain rights in the Service, branding, hosted experience, documentation, and technology, excluding Customer Content and third-party materials. These Terms do not grant a right to use Rally or Agent9 AI names, logos, or marks except to identify authorized use of the Service.
Source code or components separately made available under the Apache License 2.0 or another open-source license remain governed by that license. These Terms govern the hosted Service and do not restrict rights expressly granted by an applicable open-source license.
If you provide ideas or feedback, you grant Agent9 AI a worldwide, perpetual, irrevocable, royalty-free right to use it without restriction or compensation, provided Rally does not publicly identify you as the source without permission.
Early access, changes, and fees
Rally is in early access. Features, models, connectors, limits, interfaces, evidence formats, and availability may change. Some displayed integrations are adapter-ready or planned rather than active; the Service must not represent a provider as connected until authentication, capability discovery, and policy checks succeed.
Rally may modify, add, deprecate, or discontinue features and may impose reasonable usage, security, or spending limits. We will try to give advance notice of material changes when practical but may act immediately for security, legal, or provider reasons.
Unless a separate order states otherwise, early-access use is provided without a recurring service fee, but Customer remains responsible for fees charged by cloud, model, communications, and connected-system providers. Future paid plans will be presented before charges apply.
No service-level agreement, business associate agreement, data-processing agreement, regulated-industry addendum, professional-services commitment, or support response time applies unless separately signed.
Disclaimers, responsibility, and liability
Warranty disclaimer
TO THE MAXIMUM EXTENT PERMITTED BY LAW, THE EARLY-ACCESS SERVICE IS PROVIDED “AS IS” AND “AS AVAILABLE.” AGENT9 AI DISCLAIMS ALL EXPRESS, IMPLIED, AND STATUTORY WARRANTIES, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, NON-INFRINGEMENT, ACCURACY, AVAILABILITY, SECURITY, AND THAT OUTPUTS OR THIRD-PARTY SERVICES WILL MEET YOUR REQUIREMENTS. Some jurisdictions do not permit certain disclaimers, so they apply only to the extent allowed.
Limitation of liability
TO THE MAXIMUM EXTENT PERMITTED BY LAW, AGENT9 AI AND ITS CONTRIBUTORS WILL NOT BE LIABLE FOR INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, CONSEQUENTIAL, OR PUNITIVE DAMAGES, OR FOR LOST PROFITS, REVENUE, GOODWILL, DATA, BUSINESS OPPORTUNITY, OR COST OF SUBSTITUTE SERVICES, ARISING FROM OR RELATED TO THE SERVICE—even if advised of the possibility.
TO THE MAXIMUM EXTENT PERMITTED BY LAW, AGENT9 AI’S TOTAL LIABILITY ARISING FROM OR RELATED TO THE SERVICE WILL NOT EXCEED THE GREATER OF (A) THE AMOUNT CUSTOMER PAID AGENT9 AI FOR THE SERVICE DURING THE 12 MONTHS BEFORE THE EVENT GIVING RISE TO LIABILITY OR (B) US $100. This limit does not apply where liability cannot lawfully be limited.
Customer responsibility
Customer is responsible for its instructions, connected authority, Customer Content, provider accounts, human review, use of outputs, and compliance obligations. Customer will defend and indemnify Agent9 AI and its contributors from third-party claims arising from Customer Content, unauthorized connections, unlawful use, or material breach of these Terms, except to the extent caused by Agent9 AI’s own breach or misconduct. This obligation applies only to the extent permitted by law.
Suspension and ending use
You may stop using Rally at any time, disconnect providers, and request account deletion. Provider credentials should also be revoked directly with each provider. Agent9 AI may suspend or terminate access for a material breach, security risk, unlawful activity, nonpayment under a future paid order, provider requirement, or discontinuation of early access.
On termination, your right to use the hosted Service ends. Provisions that by their nature should survive—including ownership, confidentiality obligations, disclaimers, liability limits, dispute terms, and accrued payment obligations—will survive. Data handling after termination follows the Privacy Policy and any separate written agreement.
General terms
These Terms and any controlling written agreement are the entire agreement about the Service. You may not assign them without Agent9 AI’s consent, except with a merger or sale of substantially all relevant assets; Agent9 AI may assign them as part of a reorganization or transaction. Neither party is liable for delay caused by events beyond reasonable control.
If a provision is unenforceable, it will be modified only as needed and the remaining terms continue. Failure to enforce a provision is not a waiver. Section headings are for convenience. “Including” means “including without limitation.” Electronic notices and acceptance are valid to the extent permitted by law.
These Terms are governed by the laws applicable at Agent9 AI’s principal place of business, without regard to conflict-of-law principles. Courts with jurisdiction there will have exclusive venue, except where applicable law requires otherwise. Before filing a claim, each party will make a good-faith effort for 30 days to resolve it through written notice.
Agent9 AI may update these Terms as Rally evolves. Material changes will be identified by a new effective date and may be communicated through the Service or an account contact. Changes will not retroactively reduce rights already accrued where prohibited by law.
Legal notices and questions may be sent to:
Agent9 AI · Rally Legal
terry@agent9.dev
https://rally.agent9.dev/
