microsoft-foundry
Deploy, evaluate, fine-tune, and manage Foundry agents end-to-end with azd: hosted agent scaffold/run/deploy, prompt agent create, batch eval, continuous eval, prompt optimizer, Agent Optimizer scaffold, agent.yaml, dataset curation from traces, model fine-tuning (SFT/DPO/RFT). USE FOR: azd ai agent, azd provision/deploy, deploy agent, hosted agent, create agent, add tool to agent, invoke agent, evaluate agent, continuous eval, continuous monitoring, agent CI/CD, optimize prompt, improve prompt, optimize agent instructions, agent optimizer, deploy model, Foundry project, RBAC, role assignment,
Profile →npx skills add microsoft/azure-skills --skill microsoft-foundry --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Helps developers work with Microsoft Foundry resources, covering model discovery and deployment, complete dev lifecycle of AI agent, evaluation workflows, and troubleshooting.
How it works
- The skill prescribes a mandatory dependency check and setup step to run scripts from the skill root:
./scripts/check-and-setup-dependencies.sh(macOS/Linux) or./scripts/check-and-setup-dependencies.ps1(Windows). - It enforces reading the corresponding sub-skill document before workflow actions and requires using the Azure MCP
foundrytool for MCP-based workflows. - For azd workflows, it requires reading azd-guidance before any azd command.
- It includes multiple sub-skills (e.g., deploy, cicd, invoke, routine, observe, trace, troubleshoot, create, agent-optimizer, eval-datasets, project/resource creation, private-network, finetuning, azd-guidance) and specifies exact workflow sequences per task.
- It outlines an infrastructure and agent development lifecycle, with explicit step sequences for creating and deploying hosted Foundry agents, optimizing prompts, evaluating agents, and setting up CI/CD.
- It defines an Agent: .foundry Workspace Standard structure and guidance on agent setup, context resolution, and local cache layout.
When to use it
Use when the user intention aligns with Foundry agent lifecycle tasks: creating projects/resources, provisioning networks, deploying hosted agents, creating or updating agents, running evaluations, optimizing prompts and agent instructions, deploying models, and setting up CI/CD or monitoring.
What it can touch
The skill requires using azd commands and Foundry MCP tools through the foundry tool, reads azd-guidance, and relies on the explicit sub-skill workflows for actions like deploy, observe, agent-optimizer, eval-datasets, and more as described in the corresponding sub-skill documents.
Caveats
- It mandates mandatory steps (dependency check and azd-azd-guidance read) before proceeding with workflows.
- It discourages use for Azure Functions, App Service, or general Azure deploy/prep; those should use alternative paths (azure-deploy/azure-prepare).
- It mentions local workspace structure requirements such as
.foundry/folders and agent-metadata files for certain workflows.
# Microsoft Foundry Skill This skill helps developers work with Microsoft Foundry resources, covering model discovery and deployment, complete dev lifecycle of AI agent, evaluation workflows, and troubleshooting. ## Pre-Execution Requirements Follow each applicable subsection below before starting its corresponding action or workflow. ### Dependency Check and Setup **MANDATORY:** As the first step after this skill loads, run the dependency check and setup script below from this skill's root and wait for it to finish before continuing. The script checks first and installs only missing dependencies; it does not reinstall dependencies that are already available. **You MUST complete this check before reading or entering any sub-skill, workflow, or workflow-specific reference.** ```bash ./scripts/check-and-setup-dependencies.sh # macOS / Linux ./scripts/check-and-setup-dependencies.ps1 # Windows (pwsh) ``` Strictly follow the script output for subsequent actions. ### Workflow Guidance **MANDATORY:** Before executing ANY workflow-specific steps, you MUST read the corresponding sub-skill document. Do not call workflow-specific MCP tools for a workflow without reading its skill document. T
- Pre-Execution Requirements
- Dependency Check and Setup
- Workflow Guidance
- Foundry MCP
- azd
- Sub-Skills
- Infrastructure Lifecycle
- Agent Development Lifecycle
- Agent: .foundry Workspace Standard
- Agent: Setup References
- Agent: Common Project Context Resolution
- Step 1: Discover Agent Roots and azd Context
- Step 2: Resolve Environment and Deployment Context
- Step 3: Select Metadata Overlay and Resolve Environment
az cognitiveservices account show \
What does the microsoft-foundry skill do?
Deploy, evaluate, fine-tune, and manage Foundry agents end-to-end with azd: hosted agent scaffold/run/deploy, prompt agent create, batch eval, continuous eval, prompt optimizer, Agent Optimizer scaffold, agent.yaml, dataset curation from traces, model fine-tuning (SFT/DPO/RFT). USE FOR: azd ai agent, azd provision/deploy, deploy agent, hosted agent, create agent, add tool to agent, invoke agent, evaluate agent, continuous eval, continuous monitoring, agent CI/CD, optimize prompt, improve prompt, optimize agent instructions, agent optimizer, deploy model, Foundry project, RBAC, role assignment,
How do I install it?
Run `npx skills add microsoft/azure-skills --skill microsoft-foundry --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From microsoft/azure-skills, a repository with 1,353 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.