# AI Prompt Shield

> Enterprise-grade security layer for Generative AI applications, Large Language Models (LLMs), and autonomous agents. Prevents prompt injection exploits, blocks adversarial jailbreaks, scrubs personal health/identifiable info (PII/PHI), and audits output tokens in under 50 milliseconds.

- [Full Documentation Corpus](https://aipromptshield.in/llms-full.txt): Comprehensive detailed technical guide, SDK parameters, and complete research briefs in a single markdown payload for LLM contexts.

## Key Sections

- [Platform Architecture & Security Engine](https://aipromptshield.in/index.html#platform): Details the multi-vector defense pipeline including Semantic Vector Analysis, System Instruction Override Isolation, and the live Governance Metrics Dashboard.
- [Developer API Reference](https://aipromptshield.in/index.html#docs): Code integration parameters, official Node.js & Python SDK guidelines, error status listings, and `/api/scan.php` JSON request/response schema specifications.
- [Live Testing Console](https://aipromptshield.in/index.html#simulator): Live testing playground to evaluate threat vectors.
- [Governance Center](https://aipromptshield.in/index.html#governance): Live operational desk, custom rules configuration, and real-time logs dashboard.
- [Enterprise Offerings](https://aipromptshield.in/index.html#enterprise): Compliance specifications including SOC 2 Type II, HIPAA Compliant BAA agreements, Single Sign-On (SSO/SAML), hybrid cloud Virtual Private Cloud (VPC) options, and dedicated SLAs.
- [Pricing Tiers](https://aipromptshield.in/index.html#pricing): Monthly subscriptions structures (Developer Free, Growth Pro, Enterprise Custom Tiers) and service contract FAQs.
- [About Us & Leadership](https://aipromptshield.in/index.html#about): Our core security mission, founder details (Founder & CEO Atul K Chaudhari), leadership team members, organizational statistics, and open positions.
- [Contact Sales](https://aipromptshield.in/index.html#contact): Sales consultation pipelines, general support emails, and live demo requests scheduler.

## Technical Briefs & Research

- [Understanding Prompt Injection](https://aipromptshield.in/blog/understanding-prompt-injection.html): In-depth threat brief covering the operational mechanics of direct jailbreaks vs. indirect prompt injection exploits, outlining vectors, real-world agent scenarios, and defense policies.
- [Why Traditional Firewalls Fail](https://aipromptshield.in/blog/why-traditional-firewalls-fail.html): Outlines how syntax checking (WAF) fails against semantic attacks (such as translators, roleplay wrappers, and base64 bypasses) and maps out the necessity of semantic firewalls.
- [Achieving SOC 2 Compliance in AI](https://aipromptshield.in/blog/achieving-soc2-compliance-in-ai.html): Guidance for CTOs on satisfying Trust Services Criteria (security, privacy, processing integrity) when operating language model endpoints, focusing on log scrubbing and model validation.
- [Securing LLM Agents Against Indirect Injection](https://aipromptshield.in/blog/securing-llm-agents-indirect-injection.html): Discusses vulnerabilities in autonomous agent pipelines executing commands hidden in scanned documents, emails, or pages, with architecture mitigation guides.
- [Guide to PII Masking in Generative AI](https://aipromptshield.in/blog/guide-to-pii-masking-in-ai.html): Technical analysis comparing regular expressions, Named Entity Recognition, and Transformer models for real-time redaction of sensitive PHI and PII context elements.