v0.1.1

What's Changed

• Define LLMModel CRD types and generate manifests by @dcmcand in #17
• Add common helpers for slug generation, labels, naming by @dcmcand in #18
• Implement validating webhook for LLMModel by @dcmcand in #19
• Update CRD types: add access.public, replace raw JSON overrides by @dcmcand in #20
• Add OperatorConfig with env var parsing by @dcmcand in #21
• Implement storage reconciler (PVC, init container with lock file, OCI) by @dcmcand in #22
• Implement model service reconciler (Deployment, Service, SA, PodMonitor) by @dcmcand in #23
• Implement NetworkPolicy reconciler (default-deny, allow Gateway/EPP/Prometheus) by @dcmcand in #24
• Implement InferencePool + EPP reconciler by @dcmcand in #25
• Implement routing reconciler (AIGatewayRoutes for external + internal) by @dcmcand in #26
• Implement auth reconciler (SecurityPolicies, API key Secret, ReferenceGrant) by @dcmcand in #27
• Implement main LLMModel controller orchestrating all reconcilers by @dcmcand in #28
• Scaffold key manager with LLMModel watcher by @dcmcand in #32
• Implement API key CRUD via K8s Secrets and ConfigMaps by @dcmcand in #33
• Implement auth middleware for key manager by @dcmcand in #34
• Implement key manager HTTP API handlers by @dcmcand in #35
• Implement periodic key audit for group membership changes by @dcmcand in #36
• Add web UI, wire main.go, create Dockerfiles by @dcmcand in #37
• Add Helm chart with operator and key manager templates by @dcmcand in #38
• Add dev environment, CI workflows, and getting started guide by @dcmcand in #39
• Fix EPP crash by supplying EndpointPickerConfig by @dcmcand in #45
• Enforce model group access on internal JWT endpoint by @dcmcand in #47
• Add NebariApp route and landing-page tile for key-manager UI by @dcmcand in #46
• fix(#59): collapse pack into a single namespace by @dcmcand in #60
• fix(#61): use Keycloak JWKS path for internal SecurityPolicy by @dcmcand in #62
• fix(#57,#58): sanitize username for k8s data keys; log handler errors by @dcmcand in #63
• chart+docs: default keyManager.nebariApp.routing and add production install runbook by @dcmcand in #71
• fix(#64): disambiguate external and internal HTTPRoutes by Host header by @dcmcand in #70
• chore(release): cut v0.1.0-alpha.2 by @dcmcand in #72
• feat(alpha.3): shared-hostname routing + operator-owned TLS cert and Gateway listeners by @dcmcand in #73
• fix(alpha.4): scope per-model AIGatewayRoute to shared hostname by @dcmcand in #74
• fix(alpha.5): pin per-model AIGatewayRoute to shared-listener sectionName by @dcmcand in #75
• fix(alpha.6): narrow InferencePool selector to exclude EPP pod by @dcmcand in #76
• fix(alpha.7): pass --served-model-name to vLLM by @dcmcand in #77
• docs: rewrite install-production runbook from fresh-install validation by @dcmcand in #83
• Update copyright year and owner in LICENSE file by @viniciusdc in #97
• Enhance UI for Nebari Design by @jbouder in #94
• Add healthCheck to key-manager NebariApp landing tile (themed) by @dcmcand in #103
• Add healthCheck to key-manager NebariApp landing tile by @dcmcand in #101
• Theming by @jbouder in #102
• Add nvidia.com/gpu toleration to LLMModel requesting GPU resources by @marcelovilla in #99
• feat: BYO TLS secrets, /dev/shm for tensor parallelism, configurable rollout strategy by @tylerpotts in #93
• docs: capture k3s / on-prem GPU deltas in install-production runbook by @tylerpotts in #89
• Add PassthroughModel CRD for external provider routing and key minting by @dcmcand in #96
• chore: upgrade llm-d to release v0.7.0 by @dcmcand in #100
• chore: bump GitHub Actions to Node 24-compatible majors by @jbouder in #119
• docs: Astro/Starlight documentation site (supersedes #108) by @dcmcand in #120
• chore: opt in to the packs.nebari.dev docs portal by @dcmcand in #126
• docs: adopt @nebari/starlight shared theme plugin by @dcmcand in #127
• ci(docs): move docs workflow actions off the Node 20 runtime by @dcmcand in #128
• ci(docs): fix preview cleanup failing on unsupported per_page param by @dcmcand in #129
• ci(docs): add manual workflow to purge preview deployments by @dcmcand in #130
• docs: verify preview-cleanup lifecycle by @dcmcand in #131
• chore(docs): bump @nebari/starlight to 0.2.1 by @jbouder in #132
• docs: remove the preview-cleanup verification comment by @dcmcand in #133
• feat(dev): align local dev stack with AI Gateway v0.5 + key-manager UI dev mode by @dcmcand in #115
• React UI by @jbouder in #111

New Contributors

• @dcmcand made their first contribution in #17
• @viniciusdc made their first contribution in #97
• @jbouder made their first contribution in #94
• @marcelovilla made their first contribution in #99
• @tylerpotts made their first contribution in #93

Full Changelog: https://github.com/nebari-dev/llm-serving-pack/commits/v0.1.1