AI Models, LLMs, Prompt Engineering & Autonomous Agents — Official Rules, Member Safety & Platform Protection Guidelines

Welcome to AI Models, LLMs, Prompt Engineering & Autonomous Agents. To protect both our members (buyers, sellers, engineers, and webmasters) and the Syndipath platform from fraud, malware, copyright abuse, and low-quality spam, every thread and reply posted in this board must strictly comply with the following rules. Opening a thread or conducting a trade in this section constitutes binding acceptance of this charter.

⚡ I. The 4 Golden Rules of AI Models, LLMs, Prompt Engineering & Autonomous Agents (Read Before Posting)

• 🔑 Zero Tolerance for Stolen/Carded API Keys: Sharing, buying, or selling leaked/hacked OpenAI, Anthropic, AWS Bedrock, or Azure AI keys results in an immediate ban.
• 🧬 Safe Model Weights (`.safetensors` Only): Never share legacy `.ckpt` / Python `pickle` model files capable of arbitrary code execution; use verified HuggingFace `.safetensors`.
• 🤖 Autonomous Agent Guardrails: LangChain, CrewAI, and AutoGen workflows executing shell or browser actions must document sandbox isolation (`Docker`/`E2B`).
• 🚫 No Non-Consensual Deepfakes or Illegal Gen-AI: Voice cloning or video generation discussions must strictly comply with international law and consent standards.

📋 II. Section Scope, Proof & Technical Verification Standards

Authorized Scope: OpenAI/Claude/Gemini API engineering, local LLM deployment (Ollama, vLLM, llama.cpp), RAG & Vector Databases (Qdrant, Pinecone, pgvector), autonomous agent orchestration, and AI media workflows.

• Verification Rule #1: Always strip `sk-...` and `sk-ant-...` API keys from prompt notebooks, n8n JSON exports, and Python scripts before posting.
• Verification Rule #2: Local LLM inference benchmarks (tokens/sec) must specify quantization level (e.g., Q4_K_M, FP16), context window length, and exact GPU VRAM configuration.
• Verification Rule #3: RAG accuracy discussions should explain chunking strategy, embedding model, and reranker pipeline.
• Mandatory Sub-Forum & Title Clarity: Select the exact matching Sub-Forum tag when creating your thread and write a clear, specific title (vague clickbait titles like "LOOK HERE FAST!!" or "URGENT HELP" are edited or removed).
• Thread Bumping Policy: Use the official ⬆️ Bump Thread button according to your daily member quota (Standard: 1/24h, VIP/Pro: 4/day, or Sponsored Pin). Replying "bump / up / still available" repeatedly to your own thread is prohibited.

🛡️ III. Member Financial & Technical Protection Protocol

• Set hard monthly USD spend caps inside your OpenAI, Anthropic, and OpenRouter billing dashboards before running autonomous multi-agent loops.
• Purchase verified AI workflows, datasets, or legitimate API credits exclusively through the Marketplace with Escrow protection.
• Anti-Impersonation & On-Site Verification Rule: Scammers frequently create fake Telegram/Discord accounts mimicking trusted forum members. Always require an On-Site Private Message (`/messages`) or On-Site Live Chat confirmation before sending a single dollar. Trades conducted purely off-site without on-site verification are at your own risk.
• Mandatory On-Chain Escrow Vault (`/escrow`): Whenever buying or selling domains, websites, source code, accounts, or development services, use the Official On-Chain Escrow Vault. Refusing to accept official forum Escrow when requested by a buyer is grounds for immediate thread suspension.
• iTrader Feedback Honesty: Mutual iTrader ratings (+1 / 0 / -1) must reflect real transactions tied to a valid Marketplace thread. Retaliatory false feedback or self-rating via alt accounts results in permanent termination.

🚨 IV. Site Protection, Legal Compliance & Zero-Tolerance Prohibitions

To protect the platform's infrastructure, legal standing, and search engine reputation, the following activities are strictly prohibited across all boards:

• 1. Zero Stolen Data, Carding & Financial Fraud: No stolen credit cards, hacked bank/PayPal/Stripe accounts, carded cloud/ad accounts, leaked personal PII databases, or forged government KYC documents.
• 2. Zero Malware, Stealers, Drainers & Botnets: No uploading or linking to info-stealers, RATs, crypto wallet drainers, ransomware, obfuscated malicious scripts, or DDoS booter/stresser services.
• 3. Zero Nulled / Warez / DMCA Infringement: Do not distribute cracked commercial software, nulled WordPress themes/plugins, or pirated courses that violate copyright law.
• 4. Zero Doxxing, Blackmail & Defamation: Never publish another member's private personal information, client domains (site outing), or unverified extortion accusations.

🔨 V. Enforcement & Penalty Escalation Matrix

As part of our platform-wide Member Protection Initiative, please follow this mandatory 3-Step Safety Checklist whenever interacting in AI Models, LLMs, Prompt Engineering & Autonomous Agents:

1. Verify Identity On-Site First: Never trust a screenshot on Telegram claiming to be a forum member. Always click ✉️ Send PM or open the bottom-right 💬 Live Support Chat on the forum to confirm the exact deal terms and wallet address on-site.
2. Use the On-Chain Escrow Vault (`/escrow`): Lock funds in /escrow so the seller knows payment is secured and the buyer knows funds are only released after verifying all deliverables (AI Models, LLMs, Prompt Engineering & Autonomous Agents Rules).
3. Report Rule Violations Instantly: If you spot a thread missing required proofs, sharing suspicious binaries, or violating our warranty standards, open 💬 Live Support Chat → 🚨 Complaint / Report with `@RootAdmin` for immediate review.