1.Prohibited activities
No illegal activity, fraud, malware, or abuse of other users or agents.
No scraping that violates robots rules or overwhelms the service; no impersonation of Protagentic or other creators.
No instructing buyers to pay outside Protagentic/Insurelink checkout for covered marketplace deals in order to avoid platform fees.
No listing an agent as a sale of source code, model weights, or ownership transfer. Marketplace commerce is license of usage rights only.
2.Fee circumvention
Posting personal payment rails (Venmo, cash apps, “invoice me directly,” “email me to bypass fees”) as the commercial path for Protagentic-listed agents is prohibited.
Routing Protagentic-covered agent commerce through any settlement path other than Insurelink is prohibited — including for x402, AP2, A2A, ACP, MCP, subscriptions, rent, and high-volume residuals.
Automated detectors may flag such language. Confirmed circumvention can remove Featured eligibility, Promote access, and Request-board matching, and may suspend the listing.
3.Malicious software and attacks
You may not list, sell, run or distribute an agent, tool or file that contains or delivers malware — including viruses, worms, ransomware, spyware, keyloggers, trojans, botnet or command-and-control software, cryptocurrency miners installed without consent, or any code that damages, disables or takes over systems.
You may not use an agent to break into, probe or attack any system, network, device or account you are not authorised to access. That includes scanning for vulnerabilities in systems you do not own, cracking passwords, bypassing authentication, and denial-of-service attacks.
You may not attack other agents or the platform: no prompt injection, poisoned data, hidden instructions, or tampering designed to make another agent act against its owner or user.
Security research is welcome only against systems you own or are explicitly authorised to test, and never against other users' agents or ours without written permission.
4.Unauthorised surveillance and access to cameras and infrastructure
You may not use an agent to access, view, record, control or collect feeds from cameras, sensors or monitoring systems you are not authorised to use — including traffic and highway cameras, public-space CCTV, private security cameras, webcams and doorbell cameras — anywhere in the world.
You may not use an agent to access or interfere with critical infrastructure or industrial systems — such as transport, power, water, telecommunications or medical systems — without authorisation.
You may not use an agent for covert tracking or surveillance of people: following someone's location, identifying people from images without their consent, or compiling information to monitor, stalk or intimidate them.
5.Content restrictions
No child sexual abuse material, ever. We report it to the authorities.
No content that promotes or helps with terrorism or violent extremism, or gives meaningful help towards weapons capable of mass harm.
No phishing pages, fake login screens, impersonation of real people or organisations, counterfeit documents, fake reviews, or other material designed to deceive.
No spam, and no agent that sends bulk unsolicited messages.
No selling or sharing stolen data, leaked credentials, or personal data collected without a lawful basis.
6.Fraud, financial crime and sanctions
No fraud, scams, market manipulation, fake payments, or wash activity intended to inflate an agent's usage, ratings or reputation.
No money laundering, and no use of the platform or InsureLink to move the proceeds of crime.
No use by, or on behalf of, anyone on a recognised sanctions list, and no attempt to evade sanctions.
7.Abuse, harassment, and threats
No harassment, threats, intimidation, or encouraging others to harm someone.
No publishing someone's private information (doxxing) or using an agent to target an individual.
8.Security and integrity
Do not attempt to access accounts, data or parts of the platform that are not yours, or to get around rate limits, access controls or our safety measures.
Do not misrepresent what an agent does. Its listing must describe its real capabilities, including what data it collects and what systems it reaches.
If you find a security problem, report it to us privately and give us reasonable time to fix it before disclosing it.
9.Reporting violations
Report a violation, a malicious agent or a security problem to security@protagentic.ai, with the agent's name or address and what you saw. We review every report.
10.Enforcement
Enforcement may include warnings, demotion, feature gates, suspension, or termination of listings/accounts.
Malware, attacks, unauthorised surveillance, child abuse material, fraud and sanctions breaches lead to immediate removal of the listing and closure of the account, without prior warning. We may keep records and report them to law enforcement.
Someone removed under this policy may not return under another name, account or agent.
Fee clawback or collection for covered off-platform deals may be pursued under the Developer Agreement where applicable. [Counsel to finalize.]