CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

databricks-python-sdk

Databricks development guidance including Python SDK, Databricks Connect, CLI, and REST API. Use when working with databricks-sdk, databricks-connect, or Databricks APIs.

Quality

71%

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/databricks/claude/skills/databricks-python-sdk/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

63%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 highly actionable, code-dense reference skill whose snippets Claude could not reconstruct from prior knowledge, with strong async/wait/error patterns and a well-signaled doc-index reference. Its weaknesses are redundancy (a link table and URL-pattern explanation each duplicated), a broken reference to a nonexistent examples/ directory, and missing safeguards around destructive operations. The contradictory CLI version requirement (frontmatter '>= v1.0.0' vs body '0.278.0') is an internal consistency defect.

Suggestions

Delete the 'Quick Reference Links' table and fold the duplicated URL-pattern explanation into one place — the per-section doc links and references/doc-index.md already cover it.

Fix or remove the reference to `examples/` — the directory does not exist in the bundle; either add the worked examples or point to references/doc-index.md alone.

Reconcile the CLI version requirement: the frontmatter says 'Requires databricks CLI (>= v1.0.0)' while the body says '0.278.0 or higher'.

Add a brief pre-delete safeguard pattern (e.g., verify cluster_id via w.clusters.get before delete) to the destructive-operation examples.

DimensionReasoningScore

Conciseness

The bulk is dense, genuinely non-obvious SDK guidance (Claude cannot know these patterns), but there is real waste: the 16-row 'Quick Reference Links' table duplicates the doc link already given in every per-API section and in references/doc-index.md; the URL pattern is explained twice ('SDK Documentation Architecture' and again in 'When Uncertain'); the opening line restates the frontmatter; and unpinned version numbers ('0.278.0 or higher', contradicting the frontmatter's '>= v1.0.0') are time-sensitive. Not 2 because almost nothing explains concepts Claude already knows; not 4 because the duplicated link table and repeated URL-pattern sections are clearly trimmable.

3 / 5

Actionability

Nearly every section gives copy-paste-ready, real code: full client construction for four auth modes, per-API CRUD snippets, WRONG/CORRECT asyncio.to_thread examples, three wait patterns, and try/except error handling. Not 5 because a few snippets are not verbatim-executable — `StatementState.SUCCEEDED` is used without its import, and the pagination example imports `ListJobsRequest` but never uses it; not 3 because these are isolated minor gaps in otherwise complete, concrete guidance covering all common cases.

4 / 5

Workflow Clarity

The layout follows a sensible sequence (environment setup → configuration → usage → common patterns → a 'When Uncertain' section that prescribes fetch-and-verify before answering, a genuine validation loop), and error-handling/wait patterns supply checkpoints for fragile long-running operations. Not 5 because destructive operations (w.clusters.delete, dbutils.fs.rm recurse=True) are shown without safeguards or pre-validation steps, and setup has no verify-auth-before-first-use check; not 3 because verification guidance and error-recovery patterns are explicitly present.

4 / 5

Progressive Disclosure

The signature-table bulk is correctly split into references/doc-index.md and clearly signaled ('For a per-API table of method signatures and direct doc URLs... see references/doc-index.md'), which is anchor-4 behavior. However, the body still inlines substantial bulk reference content (the 16-row Quick Reference Links table, the URL-pattern architecture section), the referenced `examples/` directory ('Worked examples... live under examples/') does not exist in the bundle — a broken navigation path — and the assets/ files are never referenced. These place it at 'some structure but could be better organized' rather than 'minor organization gaps'.

3 / 5

Total

14

/

20

Passed

Description

70%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 solid description with an explicit and well-targeted 'Use when' clause built around exact package names. Its main weakness is the vague 'development guidance' framing, which names the domain's components but never states what the skill actually does with them. The broad 'Databricks APIs' trigger also risks overlap with the skill's own siblings.

Suggestions

Replace 'Databricks development guidance' with concrete actions, e.g., 'Create and manage Databricks clusters, jobs, SQL warehouses, and Unity Catalog resources via the Python SDK, Databricks Connect, CLI, and REST API.'

Add natural trigger variations to the when-clause: 'Use when working with the Databricks SDK (databricks-sdk), Databricks Connect (databricks-connect), the Databricks CLI, or any Databricks REST API.'

DimensionReasoningScore

Specificity

The description enumerates four specific tool areas ("Python SDK, Databricks Connect, CLI, and REST API") but the only action verb is the generic "development guidance" — no concrete operations (e.g., create clusters, run jobs, manage Unity Catalog) are named. This sits between anchor 2 ("Processes PDF files" — generic action) and anchor 4 (several specific actions); the enumerated components lift it above 2, but the absence of any concrete action keeps it below 4.

3 / 5

Completeness

Both parts are present: what ("Databricks development guidance including Python SDK, Databricks Connect, CLI, and REST API") and an explicit when with concrete trigger phrases ("Use when working with databricks-sdk, databricks-connect, or Databricks APIs"). Not 5 because the 'what' is a vague noun phrase ('guidance') rather than concrete capabilities; not 3 because the when-clause is fully explicit, not weakly implied.

4 / 5

Trigger Term Quality

The when-clause uses the literal package names users would say: "databricks-sdk, databricks-connect, or Databricks APIs" — highly natural, exact triggers. Not 5 because common variations are missing (spelled-out "Databricks SDK"/"Databricks Connect", "Databricks CLI", "Unity Catalog"); not 3 because the included terms go beyond 'some relevant keywords' — they are the precise pip/CLI names a developer would naturally mention.

4 / 5

Distinctiveness Conflict Risk

The exact package names (databricks-sdk, databricks-connect) form a clear, distinctive niche. Not 5 because the broad trigger "Databricks APIs" overlaps with the many sibling skills this skill itself lists (databricks-core, databricks-jobs, databricks-model-serving, databricks-unity-catalog), creating minor conflict risk within the same family; not 3 because the package-name triggers are far more specific than 'somewhat specific'.

4 / 5

Total

15

/

20

Passed

Validation

81%

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

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

skill_md_line_count

SKILL.md is long (636 lines); consider splitting into references/ and linking

Warning

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

relative_links

Relative link issues: 1 missing, 3 suspicious

Warning

Total

13

/

16

Passed

Repository

databricks/databricks-agent-skills

Path

plugins/databricks/claude/skills/databricks-python-sdk/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.