API Reference

Complete API documentation for pydantic-ai-shields.

Infrastructure Shields

• CostTracking — Token/USD tracking with budget enforcement
• ToolGuard — Block tools or require approval
• InputGuard — Custom input validation
• OutputGuard — Custom output validation
• AsyncGuardrail — Concurrent guardrail + LLM execution

Content Shields

• PromptInjection — Prompt injection / jailbreak detection
• PiiDetector — PII detection
• SecretRedaction — Secret/credential detection in output
• BlockedKeywords — Keyword blocking
• NoRefusals — Refusal detection

Exceptions

• GuardrailError — Base exception
• InputBlocked — Input validation failed
• OutputBlocked — Output validation failed
• ToolBlocked — Tool access denied
• BudgetExceededError — Cost budget exceeded