菜单
Automates local Azure AI Foundry model deployment workflow for only sections 9.1, 9.2, 9.3, and 9.6 of LOCAL_DEVELOPMENT.md.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | 09.deploy-ai-models-from-local |
| description | Automates local Azure AI Foundry model deployment workflow for only sections 9.1, 9.2, 9.3, and 9.6 of LOCAL_DEVELOPMENT.md. |
| argument-hint | Use this skill to list available models, validate Azure prerequisites, deploy models via Terraform, and configure API Foundry env vars. |
| compatibility | ["linux","macos","windows"] |
| license | MIT |
| user-invokable | true |
This skill follows the same wrapper-first automation pattern used in previous local setup skills.
Scope is intentionally limited to:
No other section 9 subsections are implemented by this skill.
infra/params/dev.tfvars exists (copy from tfvars.example and fill in your values if not already created).Use wrapper entrypoints. These wrappers call internal scripts and keep usage consistent.
setup-ai-models-from-local.ps1setup-ai-models-from-local.shMaintainer note:
Manage-AIModelsFromLocal.ps1manage-ai-models-from-local.shSet-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action list-models
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh list-models
List model definitions from infra/config/ai-models.json:
Set-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action list-models
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh list-models
Checks and enforces prerequisites:
az existsaz login required if missing)Microsoft.CognitiveServicesMicrosoft.AppServiceSet-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action check-prereqs -SubscriptionId "<subscription-id>"
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh check-prereqs --subscription-id "<subscription-id>"
Runs Terraform workflow in infra/:
terraform initterraform plan -var-file=<tfvars>terraform apply -var-file=<tfvars>Set-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action deploy-terraform -TfvarsPath "infra/params/dev.tfvars"
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh deploy-terraform --tfvars-path "infra/params/dev.tfvars"
Set-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action deploy-terraform -TfvarsPath "infra/params/dev.tfvars" -AutoApprove
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh deploy-terraform --tfvars-path "infra/params/dev.tfvars" --auto-approve
Updates app/backend/.env values:
FOUNDRY_ENDPOINTFOUNDRY_API_KEYAPIM_OPENAI_URL (optional)You can provide values explicitly, or infer endpoint/key from Azure using resource group + AI Services account name.
Set-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action configure-api-env `
-FoundryEndpoint "https://ai-pronghorn-dev.services.ai.azure.com/" `
-FoundryApiKey "<foundry-api-key>" `
-ApimOpenAiUrl "https://apim-pronghorn-xxx.azure-api.net/openai"
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh configure-api-env \
--foundry-endpoint "https://ai-pronghorn-dev.services.ai.azure.com/" \
--foundry-api-key "<foundry-api-key>" \
--apim-openai-url "https://apim-pronghorn-xxx.azure-api.net/openai"
Set-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action configure-api-env `
-ResourceGroupName "pronghorn-blue" `
-AIServicesName "ai-pronghorn-dev" `
-ApimOpenAiUrl "https://apim-pronghorn-xxx.azure-api.net/openai"
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh configure-api-env \
--resource-group "pronghorn-blue" \
--ai-services-name "ai-pronghorn-dev" \
--apim-openai-url "https://apim-pronghorn-xxx.azure-api.net/openai"
Run all supported subsections in order: 9.1 -> 9.2 -> 9.3 -> 9.6.
Set-Location .github/skills/09.deploy-ai-models-from-local/scripts
.\setup-ai-models-from-local.ps1 -Action all -SubscriptionId "<subscription-id>" -TfvarsPath "infra/params/dev.tfvars"
cd .github/skills/09.deploy-ai-models-from-local/scripts
bash ./setup-ai-models-from-local.sh all --subscription-id "<subscription-id>" --tfvars-path "infra/params/dev.tfvars"
After successful run:
app/backend/.env contains non-empty FOUNDRY_ENDPOINT and FOUNDRY_API_KEY.This skill will guide you through the process of setting up PostgreSQL on your local machine.
This skill will guide you through running the PostgreSQL schema migration for Pronghorn development.
This skill will guide you through setting up the local API server for Pronghorn development, including installing dependencies, configuring environment variables, and starting the server.
This skill will guide you through setting up the local frontend for Pronghorn development, including installing dependencies, configuring .env.local, and starting the Vite development server.
This skill verifies the local Pronghorn stack by checking frontend, API health endpoint, and local PostgreSQL readiness.
Automates Docker Compose full-stack workflows for local Pronghorn development, including build prerequisites, startup, verification, logs, stop, and reset operations.