CoreValue Guides

CoreValue guides are organized by user profile and use case. Find the right guide for your needs below.

Zero-Friction

Get started in seconds — no SDK install required.

• BaseURL Swap — Point your existing OpenAI SDK at CoreValue Gateway.

Getting Started

Essential guides for all profiles.

• Debugging LLM Applications — Debug with CoreValue.
• Predefined Request IDs — Manage request IDs.

SDK Guides

For developers using @cova/gateway (TypeScript) or cova-gateway (Python).

• TypeScript Quickstart — buildHeaders, wrap, fetchNative.
• Python Quickstart — Same contract, Python idioms.
• streamChat — Streaming with FinOps metadata via final SSE event.
• fetchNative — Call native-format endpoints with FinOps metadata.
• Per-Call FinOps Metadata — Parse CovaGatewayMeta in your app.
• SDK Coexistence — Four patterns for using CoreValue Gateway.

Gateway Features

For Growth and Enterprise tiers — configure gateway behavior.

• Cova-* Headers Guide — Complete request + response header directory.
• Budget Gates — Block, alert, or downgrade on spend.
• Fallback Chains — A→B→C model fallbacks.
• Disallow List — Block models or providers per org.
• Escrow — Reserve-before-execute for concurrent requests.
• Rate Limiting — Sliding window + token bucket.
• Response Caching — Cache responses to reduce cost and latency.
• Response Mapping — Anthropic/Google → OpenAI format.
• Cross-Provider Translation — OpenAI → Anthropic/Google/Bedrock.
• Prompt Expansion — Cova-Prompt-Id template resolution.
• GET /v1/models — List available models.
• Body Retention — full / metadata_only / none.
• PII Scanning — Detect and mask sensitive data.

FinOps

For FinOps Managers.

• Cost Tracking — Cost analytics and unit economics.

Integrations

For all profiles — framework and provider integration guides.

• AI Agents — Build and monitor AI agents.
• Vercel AI Gateway — Multi-model assistant.
• Vercel AI Gateway Demo — Debate simulator.
• OpenAI Structured Outputs — Chatbot with structured outputs.
• Environment Tracking — Track dev/staging/prod.
• ETL — Data extraction pipelines.
• GitHub Actions — CI/CD integration.
• OpenAI Batch API — Batch API logging.

Prompt Engineering

Generic prompt engineering techniques — applicable to any LLM.

• Prompt Engineering Overview — Technique index.
• Prompting Thinking Models — How to prompt reasoning models.