Name: microsoft-foundry
Rating: 66.4 (1 reviews)
Author: microsoft

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, permissions, quota, capacity, region, troubleshoot agent, deployment failure, AI Services, create Foundry resource, provision, knowledge index, customize deployment, onboard, availability, fine-tune, SFT, DPO, RFT, training-data, grader, distillation, fine-tuned model, large file upload. DO NOT USE FOR: Azure Functions, App Service, general Azure deploy (use azure-deploy), general Azure prep (use azure-prepare).

Quality

81%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Failed to scan

The risk profile of this skill

Quality

Content

62%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a well-structured orchestration hub with strong, validated workflow sequencing, but it loses points for repeated routing content, an incomplete executable path (delegated sub-skill files are not in the bundle), and broken progressive-disclosure references.

Suggestions

Consolidate the three overlapping routing tables (Sub-Skills, Infrastructure Lifecycle, Agent Development Lifecycle) and the redundant tip callouts into a single routing section to remove repeated project/create, deploy-model, and finetuning guidance and reduce token cost.

Resolve the broken references: either include the ~19 referenced sub-skill markdown files in the bundle, or remove/qualify the links so Claude is never directed to read non-existent documents; also link or remove the orphaned references/auth-best-practices.md.

Tighten the Common Project Context Resolution Steps 1-7 by merging the repeated source-layering guidance and the Step 5 'Effective Value / Preferred Source' table into one canonical source-of-truth table instead of restating preferred sources across steps.

Dimension	Reasoning	Score
Conciseness	The body is substantive and does not explain concepts Claude already knows, but routing is repeated across three tables (Sub-Skills, Infrastructure Lifecycle, Agent Development Lifecycle) plus redundant tip callouts that re-state the same project/create, deploy-model, and finetuning routing, so it could be tightened.	2 / 3
Actionability	Body-level guidance is concrete (named MCP tools like agent_get and prompt_optimize, specific azd variables, a copy-paste 'az cognitiveservices account show' block, ordered resolution steps), but the primary action mechanism delegates to ~19 sub-skill files that are absent from the bundle, leaving the executable path incomplete.	2 / 3
Workflow Clarity	The Common Project Context Resolution Steps 1-7 are a clearly sequenced multi-step process with explicit validation checkpoints ('verify... before using it', 'stop and ask which source is authoritative') and feedback loops, matching the 'clear sequence with explicit validation steps' anchor.	3 / 3
Progressive Disclosure	The hub is well-organized with clearly signaled one-level-deep references, but scored against the actual bundle roughly 19 referenced sub-skill files (e.g. foundry-agent/deploy/deploy.md, finetuning/SKILL.md) are missing and references/auth-best-practices.md is an unreferenced orphan, so the disclosure structure is broken in practice.	2 / 3
	Total	9 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

This is a model-quality description: third-person voice, concrete capability list, comprehensive natural trigger terms, and explicit use/do-not-use boundaries. It closely matches the rubric's good_overall_examples and earns full marks on every dimension.

Dimension	Reasoning	Score
Specificity	The description enumerates many concrete actions ("hosted agent scaffold/run/deploy, prompt agent create, batch eval, continuous eval, prompt optimizer... model fine-tuning (SFT/DPO/RFT)") rather than vague language, matching the 'lists multiple specific concrete actions' anchor.	3 / 3
Completeness	It explicitly answers what ("Deploy, evaluate, fine-tune, and manage Foundry agents..."), when ('USE FOR:'), and when-not ('DO NOT USE FOR:'), which is the textbook both-what-and-when-with-explicit-triggers pattern.	3 / 3
Trigger Term Quality	The 'USE FOR:' clause provides broad coverage of natural phrases a user would say ("deploy agent", "create agent", "optimize prompt", "troubleshoot agent", "fine-tune"), fitting the 'good coverage of natural terms' anchor.	3 / 3
Distinctiveness Conflict Risk	The Foundry-agent niche is specific and the 'DO NOT USE FOR: ... use azure-deploy / azure-prepare' clause actively prevents overlap with sibling Azure skills, matching 'clear niche with distinct triggers'.	3 / 3
	Total	12 / 12 Passed

Validation

87%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 32 missing, 30 deeper-than-1-level	Warning
referenced_paths_exist	Referenced path issues: 1 deeper-than-1-level	Warning

	Total	14 / 16 Passed

Repository: microsoft/azure-skills
Commit: 39464cb

Reviewed: about 5 hours ago

Table of Contents

Discovery Implementation Validation

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.