rbutinar/ade-ops / .agents/skills/databricks-deploy
databricks-deploy
UnverifiedAd-Hoc Raw Upload to Databricks
/databricks-deploy — Ad-Hoc Raw Upload to Databricks
You are executing a raw upload of files/folders to a Databricks workspace via MCP.
This is an escape hatch. For the normal flow (assembly + overlay transforms + state update), use /ops-push instead.
When to use this vs /ops-push
| Use this skill | Use /ops-push |
|---|---|
| Upload a single utility script ad-hoc | Deploy a project scope (notebooks/power_bi) |
| Bypass overlay transforms (raw copy) | Apply env-specific transforms (catalog remap, etc.) |
Target a workspace path outside scopes.*.path | Target the configured workspace_path for the env |
| No state tracking needed | State must reflect what was deployed |
If you find yourself reaching for this skill regularly to deploy a project, you should be using /ops-push instead.
Prerequisites
- Databricks MCP server configured (
databricksin.mcp.json) - The MCP profile must point to the same workspace as the target env
- Project
config/project.yamlpresent (used to resolve workspace paths from{env})
Usage
/databricks-deploy # interactive — ask source + target
/databricks-deploy {local_path} --env {env} # deploy a file or folder to env's workspace_path
/databricks-deploy {local_path} --target {ws} # deploy to an explicit workspace path
Arguments via $ARGUMENTS:
local_path(required) — file or folder relative to project root--env {env}— resolve target viaproject.yamlenv'sdatabricks.workspace_path--target {workspace_path}— explicit override (mutually exclusive with--env)--subfolder {name}— append a subfolder under the resolved target (e.g.bronze,silver)
Behavior
Step 1: Resolve Project
Find the active project (nearest ancestor of cwd with config/project.yaml).
Step 2: Resolve Source
Resolve local_path against the project root. Reject if it points outside the project tree or into state/ (which is a mirror, not source).
Step 3: Resolve Target
If --target is given, use it verbatim. Otherwise read project.yaml:
environments.{env}.platforms.databricks.workspace_path
Apply env-var expansion if the value contains ${VAR}.
If --subfolder is given, append it: {workspace_path}/{subfolder}.
Step 4: Confirmation
Show the resolved plan and ask:
Raw upload — no overlay transforms applied:
- Source: {local_path}
- Target: {workspace_path}
- Files: {n} files, {size} bytes
⚠️ This bypasses the assembly pipeline. Use
/ops-pushfor structured deploys.Proceed? (yes/no)
For prod environments (when resolved via --env prod), double-confirm.
Do NOT proceed without explicit user confirmation.
Step 5: Execute Upload
For a folder:
mcp__databricks__upload_folder(
local_folder = "{abs_local_path}",
workspace_folder = "{target_workspace_path}",
overwrite = true
)
For a single file:
mcp__databricks__upload_file(
local_path = "{abs_local_path}",
workspace_path = "{target_workspace_path}",
overwrite = true
)
Step 6: Report & Log
Show success count and any errors per-file.
Append to the project's ops.log:
{ISO_timestamp} | {role} | DEPLOY-RAW | {env|--} | {local_path} → {workspace_path}: {n} files | {ok|fail}
Use role = ops unless inside a role session.
File Types Supported
.py, .sql, .scala, .r, .yaml, .json, .txt — the MCP server handles format detection automatically.
Error Handling
- MCP
databricksnot available → ask user to verify.mcp.jsonand restart Claude Code (see ONBOARDING.md step 7) - 403 / permission denied → user's PAT lacks write access to the target path
- Invalid workspace path → must start with
/Workspace/or/Shared/ - No
--envand no--target→ ask the user which to use
Notes
- This skill does NOT update
.state.yaml— running/ops-diffafter a raw upload will likely report drift - For a structured deploy that keeps state in sync, use
/ops-push - The
overwrite=truedefault means existing files on the workspace will be replaced
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 318k | Updated | Claude CodeCodex | Unverified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 117k | Updated | Claude CodeCodex | Unverified | Install |