Recommended
Safety Essentials
by Stockyard
PII redaction, content filtering, prompt injection detection, and trust policies. The baseline for production.
4 modules • 2 policies • 2 alerts
Recommended
Cost Control
by Stockyard
Rate limiting, spending caps, token budgets, and caching. Keep your LLM spend predictable.
4 modules • 3 alerts
Provider
OpenAI Quickstart
by Stockyard
Pre-configured OpenAI with model routing, caching, retry logic, and a starter prompt template.
1 provider • 3 routes • 2 modules • 1 template
Provider
Anthropic Quickstart
by Stockyard
Claude model routing with caching and an analysis-focused prompt template.
1 provider • 1 route • 2 modules • 1 template
Advanced
Multi-Provider Failover
by Stockyard
OpenAI + Anthropic with automatic failover, health checks, circuit breakers, and load balancing.
2 providers • 5 modules • 2 alerts
Advanced
Evaluation Suite
by Stockyard
DAG workflows for accuracy evaluation and hallucination detection. Systematic LLM testing.
2 workflows • 1 template
Recommended
Zero-Waste AI
by Stockyard
Aggressive cost optimization: semantic caching, token trimming, output caps, model downgrade on simple queries. Cuts LLM spend 40–70%.
5 modules
Advanced
Enterprise Ready
by Stockyard
RBAC enforcement, audit logging, PII redaction, compliance evidence, IP fencing, approval gates. SOC2-ready.
7 modules
Advanced
Agent Safety
by Stockyard
Guardrails for autonomous AI agents: boundary enforcement, hallucination detection, secret scanning, consensus checks.
6 modules
Compliance
GDPR Europe
by Stockyard
EU-compliant: PII redaction enforced, data residency routing to EU providers, consent logging, audit trails for Article 30.
4 modules
Advanced
Prompt Engineering Kit
by Stockyard
Template injection, A/B testing hooks, prompt linting for common mistakes, version tracking, and quality evaluation gates.
4 modules