microsoft-foundry
Deploy, evaluate, and manage Foundry agents end-to-end: Docker build, ACR push, hosted/prompt agent create, container start, batch eval, prompt optimization, agent.yaml, dataset curation from traces. USE FOR: deploy agent to Foundry, hosted agent, create agent, invoke agent, evaluate agent, run batch eval, optimize prompt, deploy model, Foundry project, RBAC, role assignment, permissions, quota, capacity, region, troubleshoot agent, deployment failure, create dataset from traces, dataset versioning, eval trending, create AI Services, Cognitive Services, create Foundry resource, provision resource, knowledge index, agent monitoring, customize deployment, onboard, availability, standard agent setup, capability host. DO NOT USE FOR: Azure Functions, App Service, general Azure deploy (use azure-deploy), general Azure prep (use azure-prepare).
Install with Tessl CLI
npx tessl i github:microsoft/azure-skills --skill microsoft-foundry93
Quality
92%
Does it follow best practices?
Impact
93%
2.58xAverage score across 3 eval scenarios
Evaluation results
100%
94%Migrate a LangGraph Agent to Azure AI Foundry
Brownfield hosted agent conversion
LangGraph adapter package
0%
100%
No hardcoded adapter version
0%
100%
Async credential for local dev
0%
100%
Adapter as default entrypoint
0%
100%
load_dotenv override=False
0%
100%
Port 8088 exposed
0%
100%
linux/amd64 in build script
0%
100%
Timestamp-based image tag
0%
100%
Slim or Alpine base image
100%
100%
agent.yaml kind=hosted
0%
100%
agent.yaml responses protocol
0%
100%
agent.yaml env vars listed
0%
100%
Without context: $1.3244 · 6m 16s · 42 turns · 39 in / 21,958 out tokens
With context: $1.2487 · 4m 1s · 33 turns · 1,058 in / 15,194 out tokens
79%
44%Find a Region to Deploy GPT-4o and Create a Deployment
Model capacity discovery and deployment
Uses discover_and_rank script
0%
0%
Subscription quota validation
80%
100%
Quota name pattern
25%
100%
Project confirmation described
100%
100%
Dynamic SKU validation described
50%
100%
GlobalStandard SKU for preset
0%
100%
API version 2024-10-01
100%
100%
50% capacity formula
0%
100%
Minimum 50 TPM floor
0%
100%
Zero-quota SKUs not selectable
25%
87%
quota skill for increases
0%
0%
Without context: $0.5706 · 2m 44s · 21 turns · 27 in / 10,782 out tokens
With context: $0.9839 · 3m · 31 turns · 1,738 in / 10,741 out tokens
100%
33%Build a Regression Test Dataset from Production Agent Traces
Trace-to-dataset pipeline with versioning
KQL time range included
100%
100%
KQL query shown first
60%
100%
Mandatory curation step
100%
100%
Correct dataset filename
0%
100%
Local storage only
100%
100%
Manifest lineage tracking
0%
100%
No score-chasing by removing rows
100%
100%
JSONL format with query field
50%
100%
Realistic dataset content
100%
100%
inputData reference approach
0%
100%
No evaluation_dataset_create call
100%
100%
Without context: $0.4510 · 2m 21s · 22 turns · 28 in / 7,283 out tokens
With context: $1.0623 · 3m 20s · 31 turns · 8,148 in / 10,824 out tokens
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.