databricks-ai-functions

Use Databricks built-in AI Functions (ai_classify, ai_extract, ai_summarize, ai_mask, ai_translate, ai_fix_grammar, ai_gen, ai_analyze_sentiment, ai_similarity, ai_parse_document, ai_prep_search, ai_query, ai_forecast) to add AI capabilities directly to SQL and PySpark pipelines without managing model endpoints. Also covers document parsing and building custom RAG pipelines (parse → prep_search → index → query).

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Quality

Content

92%

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

The content is excellent on conciseness, actionability, and workflow clarity — lean, executable, and sequenced with error-handling feedback loops. The one real weakness is progressive disclosure: the body repeatedly links to three reference files (1-task-functions.md, 2-ai-query.md, 3-ai-forecast.md) that do not exist in the skill bundle, breaking the navigation it advertises.

Suggestions

Add the missing referenced files (1-task-functions.md, 2-ai-query.md, 3-ai-forecast.md) to the skill bundle, or remove the links and inline the essential content — currently the body points to files that are absent, which breaks progressive disclosure.

Verify the cross-skill links to ../databricks-vector-search/SKILL.md, ../databricks-genie/SKILL.md, and ../databricks-model-serving/SKILL.md resolve, since broken references undermine the same navigation principle.

Dimension	Reasoning	Score
Conciseness	The body is dense and lean: it assumes Claude's competence (no explanation of what a PDF, RAG, or VARIANT is) and every section earns its place with executable SQL and high-signal guidance, matching the 'lean and efficient; every token earns its place' anchor.	3 / 3
Actionability	Provides multiple fully executable SQL examples (chain-enrich query, ai_mask, ai_similarity self-join, ai_forecast, ai_query with from_json, full multi-stage pipeline) plus concrete PySpark snippets and exact DBR/warehouse prerequisites, matching the 'fully executable code/commands; copy-paste ready' anchor.	3 / 3
Workflow Clarity	The Document Processing Pipeline is sequenced Stage 1→2→3 with explicit error filtering (`WHERE parsed:error_status IS NULL`), per-row error routing to a sidecar table, and guidance to check `failOnError => false` / `errorMessage`, giving clear feedback loops for these batch/destructive operations per the rubric's feedback-loop guidance.	3 / 3
Progressive Disclosure	The body is well-signaled with a Reference Files section pointing one-level-deep to 1-task-functions.md, 2-ai-query.md, and 3-ai-forecast.md, but none of those referenced bundle files actually exist in the skill directory, so the progressive disclosure is broken rather than genuinely functional, matching the 'references present but not clearly signaled / structure could be better' level rather than a 3.	2 / 3
	Total	11 / 12 Passed

Description

82%

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, specific description that names concrete functions and capabilities with good natural trigger terms and a clear niche. Its main weakness is the absence of an explicit 'Use when...' clause, which caps completeness at 2.

Suggestions

Add an explicit 'Use when...' trigger clause, e.g. 'Use when the user wants to add AI capabilities to Databricks SQL or PySpark, mentions Databricks AI Functions, or needs to build a RAG/document-parsing pipeline.'

Trim or relocate the long inline function list — listing all thirteen function names in the description adds tokens; the body table covers them, so the description could name a representative few and point to the rest.

Dimension	Reasoning	Score
Specificity	Enumerates thirteen concrete named functions (ai_classify, ai_extract, ai_summarize, etc.) and states specific actions — 'add AI capabilities directly to SQL and PySpark pipelines' and 'building custom RAG pipelines (parse → prep_search → index → query)', matching the 'lists multiple specific concrete actions' anchor.	3 / 3
Completeness	It clearly answers 'what' (the function list and capabilities) but lacks an explicit 'Use when...' trigger clause for 'when'; the closest is 'Also covers document parsing and building custom RAG pipelines', which only implies use cases, so per the guideline a missing explicit trigger caps completeness at 2.	2 / 3
Trigger Term Quality	Includes natural terms users would say — 'Databricks', 'AI Functions', 'SQL', 'PySpark', 'RAG pipelines', 'document parsing' — with strong coverage of the function names themselves, matching the 'good coverage of natural terms' anchor.	3 / 3
Distinctiveness Conflict Risk	The niche is clearly Databricks AI Functions with a distinct named function set and a specific RAG pipeline flow, making it unlikely to trigger for unrelated skills; matches the 'clear niche with distinct triggers' anchor.	3 / 3
	Total	11 / 12 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
relative_links	Relative link issues: 19 missing, 3 suspicious	Warning

	Total	15 / 16 Passed

Repository: databricks-solutions/ai-dev-kit
Commit: 1c43c21

Reviewed: 3 days 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.