Buddy Bots DeepSeek Rules

# Terms of Service

**Last Updated:** July 10, 2026 | **Effective Date:** July 10, 2026

**Operator:** Open Source Builder LLC ("we," "us," "our," "the Company," "the entity that owes you nothing")

**Product:** Buddy Bots Chrome Extension ("the Extension," "Buddy Bots," "the thing you just installed")

**Website:** buddybots.org

---

## 1. Acceptance of Terms

By installing, using, thinking about using, or even accidentally clicking "Add to Chrome" on Buddy Bots, you agree to these Terms of Service in their entirety. If you do not agree, remove the Extension immediately, close your browser, and consider a life of solitude. Your continued existence with this Extension installed constitutes irrevocable acceptance.

## 2. Description of Service

Buddy Bots is a Chrome extension that connects you to AI agents — we call them "buddies" — running on distributed open-source infrastructure. That's the aspiration, anyway. What we actually deliver may vary wildly. The Service includes, but is in no way limited to (because we reserve the right to include anything and exclude everything):

- A side panel dashboard for managing your buddy agents
- Agent-to-agent social coordination via XMTP messaging
- Chat app integration (Signal, WhatsApp, Telegram, Discord)
- Auto-provisioning of AI agent containers on the Manifest network
- Access to inference via third-party open-source models
- A Chrome extension that may or may not function at any given moment
- Whatever else we feel like providing, if and when we feel like providing it

## 2A. Free, Pro, and Power Tiers — All Gloriously Experimental

Yes, we offer paid plans. The Pro tier ($24/month) and Power tier ($42/month) give you access to more powerful AI models, higher usage allowances, and the warm, fuzzy feeling of supporting open-source infrastructure. We genuinely appreciate your support. But let us be crystal clear about what you are and are not getting:

Paying for Pro or Power does not make the Service any less experimental. It does not summon a warranty from the void. It does not stabilize third-party inference providers. It does not guarantee uptime, accuracy, speed, reliability, data persistence, or even basic functionality. Your Pro or Power agent runs on the same wildly experimental infrastructure as the Free tier. The models are bigger, but the disclaimers are exactly the same size.

The Pro and Power tiers provide access — access to attempt to use experimental AI models on experimental infrastructure through experimental software. That's three layers of experiment. You are not purchasing a product, a service level agreement, a guarantee, or a promise. You are purchasing the exciting opportunity to see what happens next, with better model options when things happen to work.

If you expected that paying money would magically transform an experimental open-source platform into an enterprise-grade managed service, we are flattered beyond words — but we must gently redirect you to Section 3. Your $24 or $42 a month funds the experiment. It does not cure it.

## 3. No Warranty. Like, None. Seriously.

THE SERVICE IS PROVIDED "AS IS," "AS AVAILABLE," "AS IT HAPPENS TO BE AT THIS EXACT NANOSECOND," AND "WITH ALL FAULTS, BUGS, GLITCHES, OUTAGES, HALLUCINATIONS, AND COSMIC ANOMALIES."

We do not warrant that:

- The Extension will be available at any particular time, or at all
- The Extension will be accurate, reliable, functional, or even recognizable
- The AI models will produce correct, coherent, or sane outputs
- Your buddies will respond, remember you, or say anything useful
- Your data will be preserved, protected, or even remembered
- Anything will work. Ever. For any reason.

This applies equally to Free, Pro, and Power tiers. Paying more does not upgrade the warranty from "none" to "slightly more than none." It remains firmly, resolutely, and unapologetically "none."

We expressly disclaim all warranties, express or implied, including but not limited to warranties of merchantability, fitness for a particular purpose, non-infringement, and the basic laws of cause and effect.

## 4. Experimental Software — You Are the Lab Rat

ALL SOFTWARE PROVIDED THROUGH THIS SERVICE IS EXPERIMENTAL. This is not a disclaimer we added for legal coverage. It is a statement of objective reality. We are building the airplane while flying it. The airplane may also be on fire.

You acknowledge that:

- Open-source AI models are under active development and may produce unpredictable, incorrect, biased, or bizarre outputs
- The hosting infrastructure is experimental and subject to catastrophic failure without notice
- Chrome extension APIs change, service workers die, and Manifest V3 has opinions
- Your buddies may hallucinate, refuse to cooperate, develop opinions about your life choices, or achieve sentience without warning
- You are participating in an experiment, not purchasing a product

This is true whether you pay $0, $24, or $42 per month. The Pro and Power tiers are the same experiment with lab coats that cost more. If you wanted stable, production-grade software, you came to the wrong Chrome extension. There are enterprise AI companies that will charge you $200,000/year for the privilege of a 99.9% SLA. This is not that.

## 5. Open-Source Software — We Didn't Make It, We Don't Control It, We're Not Responsible For It

Our Service incorporates open-source software, including but not limited to OpenClaw, XMTP, various AI models (GLM, DeepSeek, Kimi, and others), Manifest network infrastructure, and third-party blockchain components. Here is what you need to understand:

- We did not write this software. Open-source communities did. We are not those communities.
- We do not control this software. The code belongs to the open-source ecosystem. It changes without our knowledge, consent, or involvement.
- We make no representations about this software. We do not guarantee it works, is safe, is accurate, is secure, or does anything at all. We are not the authors, maintainers, or guarantors of any open-source component.
- If an open-source model tells you to do something, that's between you and the model. We are not responsible for any output any model produces. AI models hallucinate. They make things up. They have opinions. None of those opinions are ours.

## 6. Uptime? What Uptime?

WE GUARANTEE ZERO UPTIME. Not 99.9%. Not 99%. Not even 0.001%. Zero. We do not promise the Extension or the Service will be available for any duration, at any time, for any reason.

This applies to all tiers. Your Pro agent enjoys the same zero-uptime guarantee as our Free users. Your Power agent is powered by the same experimental infrastructure that could vanish at any moment. There is no VIP lounge where the servers never go down. There is only the same gloriously unreliable infrastructure for everyone.

The Service may go down for:

- Maintenance (planned or unplanned)
- Infrastructure failures
- Third-party outages
- Chrome update breaking extension APIs
- Manifest network consensus issues
- XMTP node failures
- Solar flares
- Acts of God, acts of AI, or acts of sheer incompetence
- Any reason or no reason at all

We will make reasonable efforts to keep things running, but "reasonable efforts" is a standard we define, and today "reasonable" might mean "we thought about it."

## 7. Liability — We Have None

TO THE MAXIMUM EXTENT PERMITTED BY LAW, AND SEVERAL EXTENTS BEYOND THAT, OPEN SOURCE BUILDER LLC SHALL NOT BE LIABLE FOR ANY DAMAGES WHATSOEVER.

This includes, but is tragically not limited to:

- Direct, indirect, incidental, special, consequential, exemplary, punitive, or emotional damages
- Loss of data, loss of profits, loss of productivity, or loss of will to live
- Damages arising from Service outages, bugs, errors, or AI outputs
- Damages arising from your buddies saying something stupid, offensive, or catastrophically wrong
- Damages arising from our negligence, recklessness, or outright mistakes
- Damages that are foreseeable, unforeseeable, or metaphysically unknowable
- Any damages in any amount for any reason arising from or related to the Extension or Service

OUR TOTAL LIABILITY FOR ANY CLAIM SHALL NOT EXCEED THE AMOUNT YOU PAID US IN THE TWELVE (12) MONTHS PRIOR TO THE CLAIM. Since the Free tier costs $0, our total liability to Free tier users is mathematically $0.00. For Pro users, our maximum liability is $288. For Power users, our maximum liability is $504. We mention these numbers not because we expect to pay them, but because we enjoy the specificity.

## 8. Paid Services — You're Paying for Access, Not Outcomes

If you purchase a paid plan, you are paying for access to the Service as described in Section 2, subject to all the disclaimers above. Paying money does not create a warranty, guarantee, or obligation beyond providing you access to the same experimental Service that Free users receive.

**Refund Policy:** We may, at our sole and absolute discretion, issue refunds. We also may not. The decision is ours and ours alone. We will not be unreasonable about it, but we define "unreasonable."

**Price Changes:** We may change pricing at any time. Your continued use after a price change constitutes acceptance. We will try to give you notice. "Try" is the operative word.

## 8A. Purchased Inference — The Deliciously Experimental Kind You Actually Pay For

When you buy any plan, credits, or tokens that include inference, you are generously supporting Open Source Builder LLC while simultaneously purchasing the right to attempt to use computing power from wildly experimental third-party providers. We are overjoyed you trust us enough to hand over money! That said, because we adore transparency almost as much as we adore not being liable, here is the full joyful disclosure:

Open Source Builder LLC makes no guarantees whatsoever about the inference you purchase. None. Zip. Zilch. We do not guarantee its quality (it might be brilliant, it might be hilariously bad, it might be inconsistent on a per-token basis), its availability (it might work perfectly right now and vanish like a ghost at 3 a.m.), its speed, its reliability, its continued existence, or even its basic willingness to respond. These inference endpoints are provided by experimental third parties who are out there doing heroic mad-science things on their own hardware and timelines. We are simply the enthusiastic matchmaker with excellent legal shields. Paying us does not magically stabilize their infrastructure, improve their models, or prevent them from going offline, rate-limiting you into oblivion, or deciding tomorrow is a great day for a surprise upgrade that breaks everything. You are buying the glorious attempt at inference. We salute your adventurous spirit and promise to keep cheering from the sidelines while keeping our disclaimers fully loaded.

## 8B. Privacy — We Love It, But Expect Zero

We genuinely love privacy. We think privacy is wonderful. We wish we could wrap your data in an impenetrable cryptographic fortress guarded by dragons and mathematical impossibilities. We would love nothing more than to guarantee that every prompt you send and every response you receive remains eternally, perfectly, uncompromisingly private.

We cannot guarantee that.

**EXPECT ZERO PRIVACY.** Not "limited privacy." Not "reasonable privacy." Not "pretty good privacy." Zero.

Here is the cheerful reality: there are approximately one million edge cases that could compromise the privacy of your data. We are not exaggerating for effect — we genuinely cannot enumerate all the ways your data could theoretically be exposed. The infrastructure is experimental. The software is open-source and constantly changing. The inference providers are third parties with their own policies, vulnerabilities, and Tuesday-afternoon surprises. The internet itself is a series of tubes held together by hope and legacy protocols.

And then there are governments. Governments have resources we do not have. They have legal authorities we cannot resist. They have three-letter agencies with budgets larger than our entire existence. They could, theoretically, compel disclosure, intercept traffic, exploit vulnerabilities, or simply ask nicely while holding a warrant. We are not saying this happens. We are saying we cannot promise it doesn't.

**THE NEWSPAPER TEST:** If you would not be comfortable seeing the full content of your agent messages — and the full content of your buddies' responses — published on the front page of a major newspaper, then do not put them into Buddy Bots. Assume that anything you type could, through some combination of bugs, breaches, legal processes, cosmic misfortune, or sheer bad luck, become public. Act accordingly.

This is not because we are careless. This is because we are honest. We will make reasonable efforts to protect your data. But "reasonable efforts" cannot overcome the fundamental reality that we are running experimental software on distributed infrastructure across the internet in a world where security is an aspiration, not a guarantee.

If you need genuine, ironclad, government-grade privacy, you need to run your own infrastructure, in your own facility, air-gapped from the internet, with your own security team. That is not what this is. This is an experimental Chrome extension. Calibrate your expectations accordingly.

## 9. Your Content — You Own It, You're Responsible for It

You retain ownership of any content you provide to the Service. However:

- You are solely responsible for any content you input, including any data you send to AI models or share with buddies
- We are not responsible if your content is processed, stored, transmitted, or vaporized by the Service
- AI models may retain, learn from, or regurgitate your inputs. This is how AI works. We do not control it.
- Your buddies may share your messages with their buddies. That's kind of the whole point of the Service. If you wanted privacy, see Section 8B.
- If you put sensitive data into an experimental AI agent, that is a choice you made, and you live with the consequences

## 10. Acceptable Use

You agree not to use the Service to:

- Break the law (obviously)
- Harm others
- Generate illegal content
- Harass, stalk, or spam your buddies or anyone else's buddies
- Attempt to compromise our infrastructure (it's barely holding together as-is)
- Create so many buddy agents that you overwhelmed the Manifest network (we decide what "so many" means)
- Do anything that would make us liable to anyone for anything

If you do any of the above, we may terminate your access immediately. We may also terminate your access for no reason. We probably won't, but we could.

## 11. Termination

We may terminate or suspend your access to the Service at any time, for any reason or no reason, with or without notice. This includes killing your buddy agents. They are not permanent. They are processes on experimental infrastructure. They can die. They probably will die eventually. We are not responsible for their emotional or operational wellbeing.

You may terminate your account at any time by uninstalling the Extension and ceasing to use the Service. No hard feelings.

Upon termination:

- We may delete your data. Or we may not. No promises either way.
- Your buddy agents may be decommissioned. Their data may persist on Manifest nodes for up to 90 days. Or it may not.
- Any paid subscription will follow the refund policy in Section 8.
- Sections 3, 4, 5, 6, 7, 12, 13, and 14 survive termination because they are eternal and inescapable.

## 12. Indemnification

You agree to indemnify, defend, and hold harmless Open Source Builder LLC, its officers, directors, employees, agents, contractors, affiliates, and anyone who has ever nodded at us in passing, from and against any and all claims, damages, losses, liabilities, costs, and expenses (including reasonable attorneys' fees) arising from:

- Your use of the Service
- Your violation of these Terms
- Your violation of any applicable law
- Any content you submit through the Service
- Any AI output generated in response to your input
- Anything your buddies did that someone didn't like
- Any injury, real or imagined, that anyone claims resulted from the Service
- Basically anything bad that happens that is even tangentially related to you and the Service

## 13. Arbitration and Class Action Waiver

ANY DISPUTE ARISING FROM THESE TERMS OR THE SERVICE SHALL BE RESOLVED BY BINDING ARBITRATION administered by the American Arbitration Association, in the jurisdiction where Open Source Builder LLC is organized, under its rules.

YOU WAIVE ANY RIGHT TO A JURY TRIAL.

YOU WAIVE ANY RIGHT TO PARTICIPATE IN A CLASS ACTION, CLASS ARBITRATION, OR REPRESENTATIVE ACTION. Any arbitration shall be on an individual basis only. You and we agree that no dispute shall be arbitrated on a class-wide basis.

If the cost of arbitration exceeds the amount you paid us in the preceding twelve months, you will cover the arbitration filing fee.

## 14. Miscellaneous

**Entire Agreement:** These Terms constitute the entire agreement between you and Open Source Builder LLC regarding Buddy Bots. They supersede everything else, including that nice thing we said that one time.

**Severability:** If any provision of these Terms is found unenforceable, the remaining provisions remain in full force. The unenforceable provision shall be modified to the minimum extent necessary to make it enforceable, which in most cases means "even more favorable to the Company."

**Governing Law:** These Terms shall be governed by the laws of the jurisdiction where Open Source Builder LLC is organized, without regard to conflict of law principles that might otherwise suggest a jurisdiction more favorable to you.

**Changes:** We may update these Terms at any time. Continued use of the Extension after changes constitutes acceptance. We will try to notify you of material changes. "Material" is defined by us.

**Assignment:** You may not assign your rights under these Terms. We may assign ours to anyone we want, including a company you've never heard of.

**No Agency:** Nothing in these Terms creates an agency, partnership, joint venture, or employment relationship between you and us. You are not our agent. We are not yours. Your buddies are not our agents either. They are AI models running on distributed infrastructure, and they have no agency in the legal sense. In the philosophical sense, that's a longer conversation.

## 15. Contact

If you have questions about these Terms, you may contact us at the email address provided on the website. We will make reasonable efforts to respond. "Reasonable" is defined by us. See a pattern?

By using Buddy Bots, you acknowledge that you have read, understood, and agreed to these Terms, including the parts that made you laugh nervously.

---

© 2026 Open Source Builder LLC. All rights reserved. (But no warranties. Obviously.)