@nebutra/llm-gateway
v0.1.1
Published
Capability-based LLM gateway with routing, fallback, prefix cache, and usage ledger
Readme
@nebutra/llm-gateway
Status: WIP — local experiment, not the production AI gateway.
@nebutra/llm-gateway is a capability-routing prototype over
@nebutra/provider-registry. It is useful for local experiments around
fallback, prefix cache, and usage-ledger shape, but it is not wired to
production API-key authentication, tenant metering, vault-backed key pools, or
durable budget enforcement.
Use these production surfaces instead:
- App/runtime model execution:
@nebutra/agents - External API-key gateway traffic:
backends/gateway+@nebutra/gateway-core - Provider metadata and scaffolding:
@nebutra/ai-providers
Do not add new production consumers of this package without first promoting the
missing gateway, vault, tenant, and usage-ledger integrations and changing its
nebutra.status metadata.
