Microsoft Foundry Operations and Security Guide
Paste the following prompt into your AI chat to install this skill:
Please follow https://skillhub.cn/install/skillhub.md to install @org-02qudk26/microsoft-foundry.
About this skill
Problem
Microsoft Foundry workloads often get stuck on specific operational issues: hosted-agent log anomalies, Azure OpenAI quotas and rate limits, network isolation and subnet design, RBAC and guardrail setup, model routing and PTU cost, evaluation and CI/CD release. Official documentation is spread across many pages, so engineering teams repeatedly switch between troubleshooting, architecture decisions, and security compliance. This skill organizes those topics into a searchable operations reference covering troubleshooting, best practices, decision making, architecture and design patterns, limits and quotas, security, configuration, integration and coding patterns, and deployment.
How it works and boundaries
It starts by using the category index to locate the relevant section, then reads local content by line range or linked reference file. For scenarios that need current details, it calls mcp_microsoftdocs:microsoft_docs_fetch to retrieve Microsoft Learn documentation; if that MCP is unavailable, it falls back to fetch_webpage and reads the result as Markdown. The materials also note that when metadata.generated_at is older than three months, the repository should be updated; if the documentation tool is missing, it should be installed according to the guidance. The scope is best suited to teams that are planning or running Foundry agents, Foundry Models, Azure OpenAI, or Foundry IQ, and it is not intended for general cloud issues unrelated to Foundry. For Azure Government, PTU, model retirement, billing, or region availability, live quotas and official release status should still be verified.
Use Cases
- Debug hosted-agent logs and known Azure OpenAI issues.
- Design subnets and private Foundry IQ retrieval.
- Compare PTU costs and model upgrades for Foundry.
- Ship RBAC, guardrails, and authenticated agent apps.
Best For
- Foundry platform engineers maintaining hosted-agent logs, quotas, and incident recovery
- Agent security architects planning networks, RBAC, guardrails, and data privacy
- AI platform leads deciding model routing, PTU cost, and Azure OpenAI migrations
- Agent application engineers running CI/CD release, evaluation, and red-team testing
Related Skills
For independent developers, automates Git weekly reports, prioritized bug tickets, and project health checks into shareable Markdown.
Scan Windows caches, temporary files, and junk files, show space usage and risk levels, and clean selected items to free disk space.
Deploy a WeChat Service Account backend with Hermes AI, Nginx, systemd, and an admin dashboard on an Ubuntu/Debian VM.
Covers Jenkins, GitHub, and automation-related wrap workflows for IT operations and security.