CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

azure-ai

Build and troubleshoot Azure AI Search, Speech, OpenAI, and Document Intelligence apps. WHEN: AI Search, query search, vector search, hybrid search, semantic search, speech-to-text, text-to-speech, transcribe, Speech SDK logs, session ID, result ID, OCR, convert text to speech.

Quality

79%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./plugins/azure-skills/skills/azure-ai/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

68%Weight 40%Scale 1-5

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

A well-structured index-style skill that keeps the SKILL.md body lean and pushes detail into a real, verified bundle of SDK guides and a speech diagnostics hub. Its main weakness is workflow clarity: the MCP-first routing and fallback logic is implied by section order rather than stated as an explicit sequence, and no example command invocation is given.

Suggestions

Add an explicit numbered decision workflow near the top (e.g., 1. Check if Azure MCP is enabled; 2. Prefer `azure__search`/`azure__speech` commands; 3. If not enabled, ask the user to run `/mcp`; 4. For programmatic work, use the SDK quick references) instead of leaving the routing order implicit in section placement.

Include one copy-paste-ready example invocation with arguments (e.g., `azure__search` with command `search_query` and a sample query payload, or a `search_query` call for a hybrid search) so the most common task is fully executable from the body.

Trim the capability tables that restate well-known concepts ("Linguistic analysis, stemming", "Neural voices, SSML support") and consolidate the duplicate Speech diagnostics link (lines 53 and 73) into one clearly-labeled entry; also mention the log-analysis scripts or auth-best-practices.md at the top level so they are discoverable without reading diagnostics.md first.

DimensionReasoningScore

Conciseness

The body is lean — service tables, MCP command names, and links with almost no prose padding — but capability rows restate concepts Claude already knows ("Linguistic analysis, stemming", "Neural voices, SSML support") and the Speech diagnostics link is duplicated (lines 53 and 73). Efficient with minor trims possible, matching anchor 4 rather than the every-token-earns-its-place anchor 5.

4 / 5

Actionability

Concrete, executable guidance is present: named MCP tools and commands ("`azure__search` with command `search_query`", "`azure__speech` with command `speech_transcribe`") plus an explicit fallback ("ask the user to run `/mcp` and enable Azure MCP"). It falls short of anchor 5 because no example invocation with arguments is shown, leaving minor gaps in how to execute a query or transcription end-to-end.

4 / 5

Workflow Clarity

A routing path exists implicitly (MCP section marked "Preferred", then the "If Azure MCP is not enabled" fallback), but the body never states an explicit decision sequence (e.g., check MCP availability → use tool → fall back to SDK/CLI), and no end-to-end steps for common tasks like running a search query. The sequence is present but implicit with no checkpoints, matching anchor 3 rather than anchor 4's explicit clear sequence.

3 / 5

Progressive Disclosure

The body is a clean overview whose 15 direct references (14 SDK guides and speech/diagnostics.md) all exist on disk and are one level deep with clear labels. Minor gaps keep it from anchor 5: the speech sub-bundle goes a second level (diagnostics.md links to 7 sibling files), the three scripts are only discoverable through diagnostics.md, and auth-best-practices.md is not surfaced at the top level.

4 / 5

Total

15

/

20

Passed

Description

83%Weight 40%Scale 1-5

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

A strong description that explicitly states both what the skill does and when to use it, with a rich set of natural trigger phrases tied to a clearly Azure-scoped niche. The only weaknesses are minor: a few missing synonyms/variations in the trigger list and some generic trigger terms that slightly raise overlap risk with non-Azure skills.

DimensionReasoningScore

Specificity

"Build and troubleshoot Azure AI Search, Speech, OpenAI, and Document Intelligence apps" names the domain, and the WHEN clause enumerates concrete capabilities ("vector search, hybrid search, semantic search, speech-to-text, text-to-speech, transcribe, OCR, convert text to speech"). Several specific actions with minor gaps — OpenAI and Document Intelligence lack per-service actions and the leading verbs are generic — so anchor 4 fits better than the comprehensive anchor 5 or the 1-2-action anchor 3.

4 / 5

Completeness

It explicitly answers both parts: what ("Build and troubleshoot Azure AI Search, Speech, OpenAI, and Document Intelligence apps") and when ("WHEN: AI Search, query search, vector search...") with a list of concrete trigger phrases — matching the anchor-5 example structure exactly. Not 4, because the when-clause is fully explicit rather than needing more specificity.

5 / 5

Trigger Term Quality

The WHEN clause lists natural phrases users would actually say — "vector search", "speech-to-text", "transcribe", "OCR", "convert text to speech", "Speech SDK logs", "session ID" — giving good keyword coverage. It stops short of anchor 5 because common variations and synonyms are missing (e.g., "TTS", "full-text search", "embeddings", "form extraction"), and "text-to-speech" and "convert text to speech" are redundant.

4 / 5

Distinctiveness Conflict Risk

The Azure framing ("Azure AI Search, Speech, OpenAI, and Document Intelligence", "Speech SDK logs") establishes a clear niche with mostly distinct triggers. It is not anchor 5 because several generic triggers — "OCR", "transcribe", "speech-to-text", "session ID" — could also fire for non-Azure speech or OCR skills, creating minor overlap risk with closely related skills.

4 / 5

Total

17

/

20

Passed

Validation

87%

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

CriteriaDescriptionResult

relative_links

Relative link issues: 16 deeper-than-1-level

Warning

referenced_paths_exist

Referenced path issues: 16 deeper-than-1-level

Warning

Total

14

/

16

Passed

Repository

microsoft/GitHub-Copilot-for-Azure

Path

plugins/azure-skills/skills/azure-ai/SKILL.md

Reviewed

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.