# PromptGuard — Full Documentation Index

> Production-grade AI security firewall. Protect LLM applications from prompt injection, data leaks, jailbreaks, and PII exposure.

PromptGuard sits between your application and LLM providers, scanning every request and response through a six-layer detection pipeline (normalization, regex, ML ensemble, content safety, multi-turn drift, policy evaluation) with <200ms P95 latency.

## Quick Start

- Sign up: https://app.promptguard.co
- Documentation: https://docs.promptguard.co
- API Reference: https://docs.promptguard.co/api-reference/introduction
- Python SDK: pip install promptguard-sdk
- Node.js SDK: npm install promptguard
- CLI: curl -fsSL https://raw.githubusercontent.com/acebot712/promptguard-cli/main/install.sh | sh

## Integration Methods

### Auto-Instrumentation (Recommended)
```python
import promptguard
promptguard.init(api_key="pg_live_xxx")
# All LLM calls via openai, anthropic, google-generativeai, cohere, boto3 are now secured
```

### HTTP Proxy (Drop-in)
Point your LLM SDK base URL to https://api.promptguard.co/api/v1
Pass your PromptGuard key via X-API-Key header. Your provider key stays in Authorization.

### Guard API (Standalone)
POST /api/v1/guard with messages array and direction (input/output).
Returns decision (allow/block/redact), confidence, threat_type, and optional redacted_messages.

## Security Capabilities

14 specialized detectors:
- Prompt Injection: ML ensemble (F1=0.887) + regex patterns
- Jailbreak: LLM-powered analysis across 7 attack categories
- PII Detection: 39+ entity types with checksum validation
- Content Safety: LLM-based harmful intent classification
- Multi-Turn Drift: Semantic drift analysis for crescendo attacks
- Data Exfiltration: URL/encoding pattern detection
- API Key Detection: Entropy analysis + known prefix matching
- Tool Injection: Validates tool calls in agentic workflows
- Toxicity: ML-based content moderation
- Fraud Detection: Financial/social engineering patterns
- Malware: Code injection pattern detection
- URL Filtering: Domain blocklist/allowlist
- Custom Rules: Natural-language business policies
- LLM Guard: Topic filtering and entity blocklists

## Policy Presets

- default: Balanced security for general AI applications
- support_bot:strict: Optimized for customer support chatbots
- code_assistant: Enhanced protection for coding tools
- rag_system: Maximum security for document-based AI
- data_analysis: Strict PII protection for data processing
- creative_writing: Nuanced content filtering

## Proxy Configuration

Declarative retry, fallback, timeout, and circuit breaker:
```json
{
  "retry": { "max_attempts": 3, "delay_ms": 200, "backoff_multiplier": 2.0 },
  "timeout": { "request_timeout_s": 30, "connect_timeout_s": 5 },
  "fallback": { "strategy": "failover", "providers": ["openai", "anthropic", "groq"] },
  "circuit_breaker": { "enabled": true, "threshold": 5, "recovery_timeout_s": 30 }
}
```

## Supported Providers

OpenAI, Anthropic, Google Gemini, Mistral, Groq, Ollama, vLLM, AWS Bedrock, Azure OpenAI

## Supported Frameworks

LangChain, LangGraph, CrewAI, AutoGen, LlamaIndex, Haystack, Semantic Kernel, Vercel AI SDK, Pydantic AI, OpenClaw

## Developer Tools

- Python SDK: pip install promptguard-sdk (PyPI)
- Node.js SDK: npm install promptguard-sdk (npm)
- CLI: Cross-platform Rust binary with scan, redact, init, policy, redteam, MCP server
- VS Code Extension: Real-time security scanning in the editor
- Cursor Plugin: MCP-based integration for AI-assisted coding
- ChatGPT Plugin: Security scanning via ChatGPT interface

## API Endpoints

- POST /api/v1/guard - Scan messages for threats (input/output)
- POST /api/v1/security/scan - Standalone content scanning
- POST /api/v1/security/redact - PII redaction
- POST /api/v1/chat/completions - Proxied OpenAI-compatible completions
- POST /api/v1/messages - Proxied Anthropic-compatible messages
- GET /api/v1/models - List available models
- POST /api/v1/redteam/* - Red team testing endpoints
- GET /health - API health check
- GET /status - Component-level health

## Pricing

- Free: 10K requests/month, full ML detection, 24h log retention
- Pro ($49/mo): 100K requests/month, custom policies, email alerts, 7-day retention
- Scale ($149/mo): 1M requests/month, advanced analytics, 30-day retention, priority support
- Enterprise: Custom volume, SSO/SAML, dedicated support, SLA

## Full Documentation Index

- [Introduction](https://docs.promptguard.co)
- [Quick Start](https://docs.promptguard.co/quickstart)
- [Authentication](https://docs.promptguard.co/authentication)
- [First Request](https://docs.promptguard.co/first-request)
- [Tutorial](https://docs.promptguard.co/tutorial)
- [Node.js Guide](https://docs.promptguard.co/guides/node-js)
- [Python Guide](https://docs.promptguard.co/guides/python)
- [React/Next.js Guide](https://docs.promptguard.co/guides/react-nextjs)
- [Azure OpenAI Guide](https://docs.promptguard.co/guides/azure-openai)
- [Ollama Guide](https://docs.promptguard.co/guides/ollama)
- [vLLM Guide](https://docs.promptguard.co/guides/vllm)
- [LLM Providers](https://docs.promptguard.co/guides/llm-providers)
- [Migration from OpenAI](https://docs.promptguard.co/guides/migration-from-openai)
- [Projects](https://docs.promptguard.co/guides/projects)
- [Organizations](https://docs.promptguard.co/guides/organizations)
- [CLI](https://docs.promptguard.co/tools/cli)
- [Cursor](https://docs.promptguard.co/tools/cursor)
- [ChatGPT](https://docs.promptguard.co/tools/chatgpt)
- [MCP Server](https://docs.promptguard.co/tools/mcp)
- [VS Code](https://docs.promptguard.co/tools/vscode)
- [OpenClaw](https://docs.promptguard.co/tools/openclaw)
- [Python SDK](https://docs.promptguard.co/sdks/python)
- [Node.js SDK](https://docs.promptguard.co/sdks/node)
- [Streaming](https://docs.promptguard.co/sdks/streaming)
- [Security Overview](https://docs.promptguard.co/security/overview)
- [Policy Presets](https://docs.promptguard.co/security/policy-presets)
- [Custom Rules](https://docs.promptguard.co/security/custom-rules)
- [Threat Detection](https://docs.promptguard.co/security/threat-detection)
- [Compliance](https://docs.promptguard.co/security/compliance)
- [MCP Security](https://docs.promptguard.co/security/mcp-security)
- [Policy as Code](https://docs.promptguard.co/security/policy-as-code)
- [GitHub Scanner](https://docs.promptguard.co/integrations/github-scanner)
- [CI/CD Security Gate](https://docs.promptguard.co/integrations/ci-cd-security-gate)
- [Dashboard](https://docs.promptguard.co/monitoring/dashboard)
- [Observability](https://docs.promptguard.co/monitoring/observability)
- [Usage Tracking](https://docs.promptguard.co/monitoring/usage-tracking)
- [Spending](https://docs.promptguard.co/monitoring/spending)
- [Audit Logs](https://docs.promptguard.co/monitoring/audit-logs)
- [Webhooks](https://docs.promptguard.co/monitoring/webhooks)
- [Reliability & Status](https://docs.promptguard.co/advanced/reliability)
- [Proxy Configuration](https://docs.promptguard.co/advanced/proxy-config)
- [Error Handling](https://docs.promptguard.co/advanced/error-handling)
- [Rate Limits](https://docs.promptguard.co/advanced/rate-limits)
- [Best Practices](https://docs.promptguard.co/advanced/best-practices)
- [Troubleshooting](https://docs.promptguard.co/advanced/troubleshooting)
- [Pricing](https://docs.promptguard.co/pricing)
- [Changelog](https://docs.promptguard.co/changelog)
- [API Reference](https://docs.promptguard.co/api-reference/introduction)

## Links

- Website: https://promptguard.co
- Dashboard: https://app.promptguard.co
- Documentation: https://docs.promptguard.co
- Status: https://status.promptguard.co
- GitHub: https://github.com/acebot712
- Support: support@promptguard.co
- Blog: https://promptguard.co/blog