Community Guidelines
The content, conduct, and channel standards required for operating Workspaces and AI agents on TAR.
Last Updated: August 31, 20261. Purpose & Core Principles
TAR exists to make business messaging simple, transparent, and human-friendly. We connect everyday businesses and organizations to customer channels like WhatsApp, Telegram, and web storefronts.
These Community Guidelines set forth the rules of conduct and content standards applicable to all Workspace owners, invited staff, and automated workflows.
2. Channel Etiquette & Opt-Ins
When communicating with customers, patients, or community members:
- Explicit User Consent: You must have explicit, verifiable consent or an active transactional relationship before sending outbound messages over WhatsApp, SMS, or voice channels.
- Immediate Opt-Out Honor: You must promptly honor all opt-out or unsubscribe requests (such as "STOP", "UNSUBSCRIBE", or language equivalents).
- No Message Flooding: Automated reminders, promotional broadcasts, and update loops must remain concise, relevant, and bounded in frequency.
3. Prohibited Content & Abuse
The following activities and content are strictly prohibited across all TAR Workspaces and connected channels:
- Fraud & Deception: Phishing, identity theft, financial impersonation, deceptive investment schemes, or fake invoices.
- Harassment & Threats: Cyberbullying, intimidation, hate speech, stalking, or targeted harassment of individuals.
- Illegal & Exploitative Goods: Distributing unlicensed pharmaceuticals, counterfeit items, illegal narcotics, or content depicting child exploitation or non-consensual sexual material.
- System Exploitation: Attempting to disrupt relay infrastructure, bypass Gateway access controls, or transmit viruses, malware, or ransomware.
4. Responsible AI & Agentic Conduct
TAR integrates bounded AI models (Gemma, Groq Whisper, DeepSeek). When deploying automated agents:
- Transparency: Clearly identify automated bots so that recipients understand when they are interacting with an AI assistant.
- Human Agency & Override: Every automated conversation loop must allow an authorized human staff member to pause the agent and take over instantly with 1 tap.
- Gateway Consequential Bounds: Financial ledgers, medical prescriptions, inventory deductions, and user revocations must remain deterministically validated by Gateway code and not delegated to unchecked AI decisions.
5. Workforce & Access Governance
Workspace administrators are responsible for managing internal team safety:
- Principle of Least Privilege: Assign staff appropriate roles (Admin, Member, Guest) based strictly on operational necessity.
- Device Security: Ensure team devices run official application releases and immediately revoke stolen or decommissioned hardware from the Workspace switcher.
6. Enforcement & Account Actions
If a Workspace or connected channel violates these Guidelines:
- Warning & Corrective Action: For minor infractions, we will notify the Workspace owner to rectify the issue within 24 hours.
- Relay Throttling: Outbound channel dispatch may be restricted to prevent spam or flood damage.
- Suspension or Termination: Severe or repeated violations (such as phishing, hate speech, or abuse) will result in immediate suspension of cloud relay sync and credit wallets.
7. Reporting Violations & Contact
If you encounter spam, fraudulent activity, or abusive behavior originating from a TAR-connected channel, report it immediately to our trust & safety team at abuse@tarfw.com.