Acceptable Use Policy

The mission of Naagmani AI, operated by Bhakha Services Private Limited, is to provide high-performance, programmable, and secure AI runtime infrastructure for modern software teams.

This Acceptable Use Policy (“AUP”) protects the security, integrity, and availability of our platform, integrated AI providers, and developer ecosystem. By using the Services, you agree to comply strictly with this policy. Violations may lead to immediate suspension of API keys, termination of accounts, and where appropriate, referral to legal authorities.

You may not use Naagmani AI to facilitate, encourage, or conduct any activity that violates applicable local, national, or international laws or regulations. This includes, without limitation:

• Engaging in terrorism, violent extremism, or organized criminal enterprise;
• Facilitating the illegal trade of controlled substances, firearms, or munitions;
• Violating international trade sanctions, export controls, or anti-money laundering legislation;
• Violating statutory consumer protection, financial disclosure, or privacy laws.

You must not compromise the technical security, integrity, or stability of the platform. Prohibited actions include:

• Attempting to gain unauthorized access to any system, server, account, or private network;
• Intercepting, harvesting, or stealing API credentials, private keys, or authentication tokens;
• Distributing computer viruses, worms, Trojan horses, rootkits, or automated malware;
• Conducting port scans, penetration tests, or vulnerability stress tests without prior written authorization from our security team;
• Reverse engineering, decompiling, or disassembling proprietary runtime engines or compiled platform components.

Naagmani provides programmable routing, skills, and AI firewall capabilities. You agree not to misuse or circumvent these AI protective mechanisms, including:

• Engineering adversarial prompt injections, jailbreaks, or evasion attacks intended to bypass model safety filters or platform guardrails;
• Subverting or modifying firewall inspection rules to transmit illicit payloads to upstream AI providers;
• Using the platform to generate automated cyberattack code, exploit payloads, or malicious scripts;
• Extracting or scraping training weights, embeddings, or proprietary system prompts without authorization.

The platform must not be used to deceive, mislead, or commit fraud. Prohibited activities include:

• Deploying phishing schemes, credential theft landing pages, or social engineering lures;
• Impersonating any real person, business entity, government authority, or public official;
• Generating deceptive synthetic media (deepfakes) with malicious or defamatory intent;
• Disseminating coordinated deceptive disinformation or automated astroturfing campaigns.

You may not transmit, process, or store content through our platform that contains or promotes:

• Child Sexual Abuse Material (CSAM) or any form of child exploitation;
• Incitement to suicide, self-harm, severe bodily injury, or violence against individuals or groups;
• Unlawful hate speech promoting violence or discrimination based on protected characteristics;
• Non-consensual disclosure of personally identifiable financial, medical, or intimate data.

To preserve quality of service for all developers, you must not abuse platform resources, including:

• Circumventing organization rate limits, token quotas, concurrency caps, or project budgets;
• Creating multiple throwaway accounts or fraudulent organizations to exploit free tier allotments;
• Launching distributed denial-of-service (DDoS) attacks against our edge proxies or APIs;
• Generating continuous unthrottled requests intended to disrupt platform availability.

Because Naagmani routes requests to third-party foundation models (such as OpenAI, Anthropic, Google, DeepSeek, etc.), you agree to adhere to the applicable usage policies and terms of each upstream provider you select. Any upstream violation reported by a provider to Naagmani constitutes a direct violation of this policy.

We employ automated filters, rate-monitoring telemetry, and security triage to identify anomalous activity. When a violation is identified, we may take proportionate remedial actions depending on severity:

If you discover or suspect an abuse of our platform, security circumvention, or harmful content generation, please report the incident to our trust and safety team at: