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

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

92%

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

A lean, highly actionable skill body with executable SQL/PySpark examples and a well-sequenced, error-aware document/RAG pipeline. The one gap is progressive disclosure: the three referenced deep-dive files are not present in the bundle, so the navigation structure cannot be verified.

Suggestions

Ship the referenced bundle files (1-task-functions.md, 2-ai-query.md, 3-ai-forecast.md) alongside SKILL.md so the progressive-disclosure links resolve to real content; missing files make the structure unverifiable.

If any reference file cannot be provided, convert its section into concise inline guidance or remove the link to avoid dangling references.

Dimension	Reasoning	Score
Conciseness	The body is dense and assumes competence — it skips explaining what AI Functions or RAG are and instead delivers a function matrix, runnable SQL/PySpark patterns, and a staged pipeline with no filler or padding, so every token earns its place.	3 / 3
Actionability	Guidance is concrete and copy-paste ready: a full function matrix with signatures, executable SQL chaining examples (ai_classify…:response[0]::STRING), PySpark expr() forms, and complete CREATE OR REFRESH STREAMING TABLE pipeline stages.	3 / 3
Workflow Clarity	The document-processing and RAG pipelines are explicitly staged (Stage 1 parse → Stage 2 classify → Stage 3 extract), with error-routing feedback (result:error_message, failOnError => false) and a Common Issues table giving recovery steps — clear sequence with error-recovery guidance for batch operations.	3 / 3
Progressive Disclosure	The overview points to clearly-signaled one-level-deep references (1-task-functions.md, 2-ai-query.md, 3-ai-forecast.md), but those referenced bundle files are not present in the directory, so the progressive-disclosure structure is asserted but not verifiable against an actual bundle.	2 / 3
	Total	11 / 12 Passed

Description

77%

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 specific, well-scoped description that names concrete functions and a clear Databricks niche. Its main weakness is the absence of an explicit 'Use when…' trigger clause, which leaves the 'when to use' guidance implicit.

Suggestions

Add an explicit 'Use when…' clause naming natural triggers (e.g. 'Use when adding AI capabilities to Databricks SQL/PySpark, building a RAG pipeline, or when the user mentions ai_classify/ai_extract/ai_summarize/ai_query/etc.').

Trim the long enumerated function list inside the description — the body already lists them; keeping a shorter representative set plus the trigger clause would improve conciseness without losing specificity.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete 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) and names the concrete actions of adding AI to SQL/PySpark pipelines and building a parse→prep_search→index→query RAG pipeline.	3 / 3
Completeness	It clearly answers 'what' (use built-in AI Functions in SQL/PySpark, build RAG pipelines) but has no explicit 'Use when…' trigger clause; the 'when' is only implied by the function list, which caps completeness at 2 per the rubric guideline.	2 / 3
Trigger Term Quality	Natural terms a user would say appear throughout — 'SQL and PySpark pipelines', 'RAG pipelines', 'document parsing', 'model endpoints' — and the enumerated function names (ai_classify, ai_summarize, etc.) are exactly the strings a user would mention when asking for this capability.	3 / 3
Distinctiveness Conflict Risk	The Databricks AI Functions niche is highly specific and distinct, anchored by named functions and DBR/version context, making it very unlikely to trigger for the wrong skill.	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
Path: databricks-skills/databricks-ai-functions/SKILL.md
Commit: a7e1d51

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.