databricks-agent-bricks

Create and manage Databricks Agent Bricks: Knowledge Assistants (KA) for document Q&A, Genie Spaces for SQL exploration, and Supervisor Agents (MAS) for multi-agent orchestration. Use when building conversational AI applications on Databricks.

Quality

—

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Advisory

Suggest reviewing before use

Quality

Content

65%

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

The body is highly actionable with concrete tool parameters and a runnable example, and it avoids padding. It is held back by mild redundancy, a lack of validation checkpoints around destructive operations, and inline API reference plus dangling detail-file references.

Suggestions

Add an explicit validation/retry checkpoint for destructive and batch operations, e.g. run find_by_name to confirm the target tile_id before delete, and verify endpoint_status reaches ONLINE before adding examples or routing agents.

Move the MCP Tools parameter reference (manage_ka / manage_mas field lists) into a references/ file and keep SKILL.md as an overview that links to it, which would also resolve the missing 1-knowledge-assistants.md / 2-supervisor-agents.md detail-file references.

De-duplicate the Prerequisites, Typical Workflow step 1, and Best Practices sections, and merge the Related Skills and See Also lists, to remove repeated guidance.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean (tables, parameter lists, no explaining of basic concepts), but Prerequisites, Typical Workflow step 1, and Best Practices restate overlapping guidance, and there are both a "Related Skills" and a "See Also" section, so it could be tightened.	2 / 3
Actionability	Tool calls list concrete actions (create_or_update, get, find_by_name, delete) with required/optional parameters, and the Example block gives a copy-paste-ready manage_mas call with all agent fields populated, matching the fully-executable anchor.	3 / 3
Workflow Clarity	The Typical Workflow provides a 4-step sequence with a provisioning status check, but destructive actions (delete) and batch brick creation have no validate-then-proceed feedback loop, capping this at 2 per the destructive-operations guideline.	2 / 3
Progressive Disclosure	Sections are organized and one-level-deep detail files are signaled in See Also, but those referenced files (1-knowledge-assistants.md, 2-supervisor-agents.md) are not present in the bundle, and the MCP Tools section is a large inline API reference that would better live in a separate file.	2 / 3
	Total	9 / 12 Passed

Description

85%

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: it names concrete capabilities across three brick types and includes an explicit "Use when" trigger with a clear Databricks niche. The only weakness is reliance on product jargon over the more natural terms a user might actually say.

Dimension	Reasoning	Score
Specificity	"Create and manage Databricks Agent Bricks" then enumerates three concrete component types with their purpose ("Knowledge Assistants (KA) for document Q&A", "Genie Spaces for SQL exploration", "Supervisor Agents (MAS) for multi-agent orchestration"), matching the multiple-specific-actions anchor.	3 / 3
Completeness	It states what the skill does (create/manage three brick types) and gives an explicit trigger ("Use when building conversational AI applications on Databricks"), satisfying the both-what-and-when anchor.	3 / 3
Trigger Term Quality	It includes some natural phrases ("document Q&A", "SQL exploration", "conversational AI applications") but leans heavily on product-specific jargon ("Agent Bricks", "Genie Spaces", "Supervisor Agents") and misses common user variations, fitting the some-keywords-but-missing-variants anchor rather than full coverage.	2 / 3
Distinctiveness Conflict Risk	The Databricks Agent Bricks framing plus the explicit Databricks-scoped trigger mark a clear niche unlikely to fire for unrelated skills, matching the clear-niche 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: 6 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.