TaskHound
Features Pricing About
Sign in Start your team
Features Pricing About Sign in →

Legal documents

Terms of Service Acceptable Use Policy Privacy Policy Service Level Agreement

Questions about any of these documents? Email hello@taskhound.com.

TaskHound Service Level Agreement

Effective Date: July 2, 2026  |  Last updated: July 2, 2026

TL;DR - Key points in plain English

  • 99.9% availability: we commit to keeping your TaskHound Container and the TaskHound control plane available 99.9% of each month.
  • What counts: your Container or the dashboard being unreachable due to our fault.
  • What doesn't: AI provider (Anthropic/OpenAI/Google/xAI) outages, your own workloads or misconfigurations, th.works demo containers you deployed, and announced maintenance.
  • Remedy: service credits on future invoices - 5% for 15+ minutes of downtime, up to 60% for 48+ hours, capped at one month's fee.
  • How to claim: email support@taskhound.com within 7 days of the incident with the details.

This Service Level Agreement ("SLA") sets forth the availability commitments and remedies provided by Snaju LLC ("Snaju," "we," "us") to customers of TaskHound ("Customer," "you"). This SLA is incorporated into and subject to the TaskHound Terms of Service.

Article I - Definitions

Section 1.1 - Definitions.

  1. "Covered Service" means (i) the Customer's TaskHound Container and (ii) the TaskHound control plane (dashboard, APIs, and session streaming) for paid subscriptions.
  2. "Downtime" means a period during which a Covered Service is unavailable or unreachable due to causes attributable to Snaju, as verified by Snaju's monitoring or reasonable investigation.
  3. "Monthly Availability Percentage" means the total minutes in a calendar month minus Downtime minutes, divided by total minutes in the month, multiplied by 100.
  4. "Scheduled Maintenance" means planned maintenance for which Snaju provides advance notice.
  5. "Service Credit" means the percentage of the affected month's subscription fee credited to Customer's account as the remedy for Downtime.

Article II - Availability Commitment

Section 2.1 - Commitment. Snaju commits to a Monthly Availability Percentage of at least ninety-nine point nine percent (99.9%) for Covered Services, excluding Scheduled Maintenance and the exclusions in Article VI.

Section 2.2 - What Constitutes Downtime. Downtime occurs when: (a) the Customer's TaskHound Container is unreachable or unable to run sessions; or (b) the TaskHound control plane is unavailable such that Customer cannot access sessions, the board, or its data - in each case due to failure of Snaju-managed systems or infrastructure, and not due to an exclusion.

Article III - Service Credits

Section 3.1 - Credit Schedule. If Snaju fails to meet the commitment in Section 2.1, Customer is entitled to Service Credits against the affected Team's monthly subscription fee:

Cumulative Downtime in a Month Service Credit (% of monthly fee)
15 minutes – 59 minutes5%
1 hour – 2 hours 59 minutes10%
3 hours – 8 hours 59 minutes20%
9 hours – 23 hours 59 minutes30%
24 hours – 47 hours 59 minutes40%
48 hours or more60%

Section 3.2 - Maximum. Total Service Credits for a calendar month shall not exceed one hundred percent (100%) of that month's subscription fee for the affected Team.

Section 3.3 - Application. Service Credits are applied against future TaskHound invoices; they are not transferable, not redeemable for cash, expire twelve (12) months after issuance if unused, and are forfeited on account termination.

Article IV - Claim Procedure

Section 4.1 - Deadline. Claims must be submitted within seven (7) days after the Downtime incident concludes.

Section 4.2 - Requirements. Email support@taskhound.com with subject "SLA Claim", including: the Team name, the dates and times the Downtime began and ended (with time zone), a description of the impact, and any supporting evidence.

Section 4.3 - Review and Issuance. Snaju will review claims within five (5) business days and apply approved Service Credits within ten (10) business days of approval. Snaju's determination of Downtime duration and cause, based on its monitoring data, is final.

Article V - Covered Events

Section 5.1 - Qualifying Events. This SLA covers Downtime resulting from failure of Snaju-managed hardware or host software, failure of the TaskHound control plane, failure of Snaju-managed network infrastructure, data-center power failure affecting Snaju infrastructure, and errors by Snaju personnel in managing infrastructure.

Article VI - Exclusions

Section 6.1 - Excluded Events. No Service Credits are issued for unavailability or degradation arising from:

  1. Agent Provider issues: outages, rate limits, degradations, model changes, or account actions of Anthropic, OpenAI, Google, xAI, or any other Agent Provider;
  2. Customer workloads: anything Customer or its agents run inside the Container - including deployments on th.works, resource exhaustion of the subscribed Container tier, or misconfiguration of Customer software;
  3. Scheduled Maintenance announced at least forty-eight (48) hours in advance, and emergency maintenance reasonably required for security or stability;
  4. Suspension for non-payment or enforcement under the Terms of Service or AUP;
  5. Third-party services not controlled by Snaju, including Stripe, app-store distribution, push-notification networks, and Customer's own origin git hosts;
  6. External attacks targeting Customer's deployments, unless Snaju's mitigation causes extended outage;
  7. Force majeure events beyond Snaju's reasonable control;
  8. Upstream network failures outside Snaju's reasonable control, except where Snaju fails to implement reasonable redundancy;
  9. Beta, preview, or free services, including trial periods.

Article VII - Scheduled Maintenance

Section 7.1 - Windows and Notice. Routine maintenance is performed with at least forty-eight (48) hours notice by email, dashboard notice, or status page; emergency maintenance may occur with as much notice as reasonably practicable. Snaju schedules impactful maintenance in low-usage windows wherever possible.

Article VIII - Exclusive Remedy

Section 8.1 - Exclusive Remedy. SERVICE CREDITS ARE CUSTOMER'S SOLE AND EXCLUSIVE REMEDY FOR ANY FAILURE BY SNAJU TO MEET THE AVAILABILITY COMMITMENT. This SLA does not create warranties beyond those expressly stated and does not modify the limitation-of-liability provisions of the Terms of Service.

Section 8.2 - Aggregation. Multiple Downtime incidents in a month are aggregated for the credit schedule, subject to the cap in Section 3.2.

Article IX - Claim Denial

Section 9.1 - Grounds. Snaju may deny claims that are submitted after the deadline, fall within an exclusion, are made while the account is not in good standing, are unsupported by evidence, or are determined by investigation not to be attributable to Snaju's infrastructure.

Article X - Modifications

Section 10.1 - Changes. Snaju may modify this SLA by posting an updated version with a revised "Last updated" date. Material changes adverse to Customer take effect thirty (30) days after posting. Continued use after the effective date constitutes acceptance.

Article XI - Contact

Snaju LLC - TaskHound
Support: support@taskhound.com
General: hello@taskhound.com
League City, Texas, United States

BY USING TASKHOUND, YOU ACKNOWLEDGE THAT YOU HAVE READ AND UNDERSTOOD THIS SERVICE LEVEL AGREEMENT AND ACCEPT ITS TERMS.

Put the hound on the trail.

One AI Desk, every agent, the whole team watching. Your TaskHound Container ignites in minutes.

Start your team - $19.95/user/mo Tour the features
TaskHound

The AI Desk - an orchestration platform for team AI environments. Your whole team, one shared TaskHound Container, every agent in the open.

Product

  • Features
  • TaskHound Pipeline
  • Mobile & desktop apps
  • Pricing
  • How it works
  • FAQ

Company

  • About Snaju
  • hello@taskhound.com
  • Sign in

Legal

  • Terms of Service
  • Acceptable Use Policy
  • Privacy Policy
  • Service Level Agreement
© 2026 Snaju LLC. All rights reserved.

Built by Snaju - infrastructure with flight heritage on the Moon and the ISS.

Runs on the DartNode network