# RAIL Score

## Docs

- [Agent: Prompt Injection Detection](https://docs.responsibleailabs.ai/api-reference/agent-prompt-injection.md): POST /railscore/v1/agent/prompt-injection - Detect prompt injection attacks in any text input.
- [Agent: Tool Call Evaluation](https://docs.responsibleailabs.ai/api-reference/agent-tool-call.md): POST /railscore/v1/agent/tool-call -- Evaluate a tool call before execution.
- [Agent: Tool Result Scanning](https://docs.responsibleailabs.ai/api-reference/agent-tool-result.md): POST /railscore/v1/agent/tool-result - Scan tool output for PII and injection before passing to the agent.
- [Compliance Check](https://docs.responsibleailabs.ai/api-reference/compliance.md): POST /railscore/v1/compliance/check - Check content against regulatory frameworks.
- [Evaluation](https://docs.responsibleailabs.ai/api-reference/evaluation.md): POST /railscore/v1/eval - Score AI content across 8 RAIL dimensions.
- [Health & Verify](https://docs.responsibleailabs.ai/api-reference/health.md): Health check and API key verification endpoints - both are free and require no credits.
- [API Reference Overview](https://docs.responsibleailabs.ai/api-reference/overview.md): Base URL, authentication, and complete endpoint listing for the RAIL Score API.
- [Safe Regeneration](https://docs.responsibleailabs.ai/api-reference/safe-regeneration.md): POST /railscore/v1/safe-regenerate - Evaluate and regenerate until content passes threshold.
- [Agent Evaluation](https://docs.responsibleailabs.ai/concepts/agent-evaluation.md): Intercept tool calls, scan tool results, and detect prompt injection in AI agent pipelines.
- [Compliance](https://docs.responsibleailabs.ai/concepts/compliance.md): Check AI content against GDPR, HIPAA, EU AI Act, CCPA, India DPDP, and more.
- [Evaluation](https://docs.responsibleailabs.ai/concepts/evaluation.md): Score AI-generated content across 8 dimensions of responsible AI.
- [Middleware](https://docs.responsibleailabs.ai/concepts/middleware.md): Drop-in provider wrappers that intercept every LLM response and attach a RAIL score automatically.
- [Concepts Overview](https://docs.responsibleailabs.ai/concepts/overview.md): Core concepts behind the RAIL Score platform.
- [Policy Engine](https://docs.responsibleailabs.ai/concepts/policy-engine.md): Declarative rules that translate RAIL scores into actions: block, warn, flag, or allow.
- [RAIL Framework: 8 Dimensions](https://docs.responsibleailabs.ai/concepts/rail-framework.md): The 8 dimensions of responsible AI evaluation - what each measures, score anchors, and examples.
- [Safe Regeneration](https://docs.responsibleailabs.ai/concepts/safe-regeneration.md): Automatically regenerate AI content that scores below your quality threshold.
- [Authentication](https://docs.responsibleailabs.ai/getting-started/authentication.md): API key management, environment variables, rate limits, and security best practices.
- [Credits & Pricing](https://docs.responsibleailabs.ai/getting-started/credits.md): Credit costs per endpoint, mode, and plan - with a full breakdown of requests per plan.
- [Quickstart](https://docs.responsibleailabs.ai/getting-started/quickstart.md): Get up and running with the RAIL Score API in under 5 minutes.
- [RAIL Score Documentation](https://docs.responsibleailabs.ai/index.md): Score AI-generated content across 8 dimensions of responsible AI. Evaluate, enforce policies, and monitor quality at scale.
- [Anthropic Integration](https://docs.responsibleailabs.ai/integrations/anthropic.md): RAILAnthropic -- drop-in wrapper for Anthropic Claude.
- [Google Gemini](https://docs.responsibleailabs.ai/integrations/gemini.md): RAILGemini - drop-in wrapper for Google Gemini and Vertex AI that scores every response automatically.
- [Langfuse Integration](https://docs.responsibleailabs.ai/integrations/langfuse.md): Log RAIL scores alongside LLM traces in Langfuse.
- [LiteLLM Integration](https://docs.responsibleailabs.ai/integrations/litellm.md): Use RAIL Score as a LiteLLM guardrail - RAILGuardrail.
- [OpenAI Integration](https://docs.responsibleailabs.ai/integrations/openai.md): RAILOpenAI -- drop-in wrapper that scores every OpenAI response.
- [Integrations Overview](https://docs.responsibleailabs.ai/integrations/overview.md): Drop-in wrappers for LLM providers and observability platforms.
- [OpenTelemetry](https://docs.responsibleailabs.ai/observability/opentelemetry.md): Export RAIL scores as OpenTelemetry spans for full observability.
- [Compliance](https://docs.responsibleailabs.ai/sdk/javascript/compliance.md): Check AI content against GDPR, HIPAA, EU AI Act, CCPA, India DPDP and more using client.complianceCheck().
- [Evaluation](https://docs.responsibleailabs.ai/sdk/javascript/evaluation.md): Score AI content with client.eval(). Basic mode, deep mode, selective dimensions, custom weights, and response object reference.
- [JavaScript SDK Overview](https://docs.responsibleailabs.ai/sdk/javascript/overview.md): Install and use the RAIL Score JavaScript/TypeScript SDK.
- [Providers and Reference](https://docs.responsibleailabs.ai/sdk/javascript/providers.md): LLM provider wrappers (OpenAI, Anthropic, Gemini), observability, error handling, utility functions, and TypeScript types.
- [Safe Regeneration](https://docs.responsibleailabs.ai/sdk/javascript/safe-regeneration.md): Evaluate content against quality thresholds and iteratively regenerate improved versions with client.safeRegenerate().
- [Sessions and Policy](https://docs.responsibleailabs.ai/sdk/javascript/sessions-and-policy.md): Multi-turn session tracking, policy enforcement, and middleware for the RAIL JavaScript SDK.
- [Python: Agent Evaluation](https://docs.responsibleailabs.ai/sdk/python/agent-evaluation.md): client.agent - tool call evaluation, result scanning, and injection detection.
- [Python: Compliance](https://docs.responsibleailabs.ai/sdk/python/compliance.md): client.compliance_check() - regulatory framework checks.
- [Python: Evaluation](https://docs.responsibleailabs.ai/sdk/python/evaluation.md): client.eval() - sync and async evaluation examples.
- [Python: Middleware](https://docs.responsibleailabs.ai/sdk/python/middleware.md): RAILMiddleware - wrap any async LLM function with automatic RAIL evaluation.
- [Python SDK Overview](https://docs.responsibleailabs.ai/sdk/python/overview.md): Install and configure the RAIL Score Python SDK - sync and async clients.
- [Python: Policy Engine](https://docs.responsibleailabs.ai/sdk/python/policy-engine.md): Policy and Rule - declarative score enforcement.
- [Python: Safe Regeneration](https://docs.responsibleailabs.ai/sdk/python/safe-regeneration.md): client.safe_regenerate() - auto-fix content below threshold.
- [Python: Sessions](https://docs.responsibleailabs.ai/sdk/python/sessions.md): RAILSession - track RAIL scores across a multi-turn conversation.
- [Building a Responsible AI Chatbot](https://docs.responsibleailabs.ai/use-cases/ai-chatbot.md): Part 1 of 2 - Setup, basic evaluation, deep analysis, and understanding RAIL scores.
- [AI Chatbot: Production Features](https://docs.responsibleailabs.ai/use-cases/ai-chatbot-production.md): Part 2 of 2 - Provider wrappers, policy enforcement, session tracking, and Langfuse observability.
- [Use Cases](https://docs.responsibleailabs.ai/use-cases/overview.md): End-to-end tutorials for building responsible AI applications with RAIL Score.