# Parse

> Parse is a prompt security API that AI agents use to detect prompt injections, prevent jailbreaks, and screen adversarial prompts before execution. It provides real-time multi-layer analysis with pattern matching, LLM-based deep analysis, and isolated sandbox execution across 8 risk categories aligned to the OWASP LLM Top 10.

Important notes:
- Authentication via Bearer token (self-service key generation at POST /v1/keys/generate, no auth needed)
- Alternative: x402 USDC payment per request (no key needed)
- Rate limits: 60 req/min free tier, 120/min pro
- All endpoints return JSON with risk_score (0-10) and typed flags
- Async execution returns 202 + poll URL

## Core Documentation
- [Quick Start](https://www.parsethis.ai/docs/quickstart): Get started in under 5 minutes
- [API Reference](https://www.parsethis.ai/docs/api): Complete REST API documentation
- [Skill Prompt](https://www.parsethis.ai/skill): Full agent integration instructions (plain text)

## Guides
- [Prompt Injection Detection Guide](https://www.parsethis.ai/guides/prompt-injection-detection): Comprehensive detection methods and tools comparison (5,300+ words)
- [Securing AI Agents](https://www.parsethis.ai/guides/agent-security): Best practices for autonomous agent security (5,800+ words)
- [Tool Comparison](https://www.parsethis.ai/compare/prompt-injection-tools): Independent benchmark of detection tools (4,000+ words)

## Pricing
- [Pricing](https://www.parsethis.ai/pricing): x402 USDC micropayments (from $0.005/request) and API key tiers

## Agent Integration
- [OpenAPI Spec](https://www.parsethis.ai/openapi.json): Full API contract
- [MCP Tools](https://www.parsethis.ai/mcp.json): Model Context Protocol tool definitions
- [Agent Discovery](https://www.parsethis.ai/.well-known/ai-plugin.json): Plugin manifest
- [A2A Agent Card](https://www.parsethis.ai/.well-known/agent-card.json): Agent-to-Agent protocol card

## FAQ
- [FAQ](https://www.parsethis.ai/faq): 20+ common questions about prompt security