Megha21-19/claude-code-data-eng-kit
Minimal Claude Code setup for a Databricks medallion lakehouse — skills, sub-agents, hooks, MCP server, settings.
ARCHETYPE
Pragmatist
A balanced, no-drama setup: some rules, some tools, nothing extreme.
Copy this rig
# review before running: this installs third-party code
$ npx degit Megha21-19/claude-code-data-eng-kit/.claude ./rig-claude-code-data-eng-kit # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit Megha21-19/claude-code-data-eng-kit/.claude/skills/dq-check .claude/skills/dq-check $ npx degit Megha21-19/claude-code-data-eng-kit/.claude/skills/new-bronze-ingest .claude/skills/new-bronze-ingest $ npx degit Megha21-19/claude-code-data-eng-kit/.claude/skills/promote-to-silver .claude/skills/promote-to-silver $ npx degit Megha21-19/claude-code-data-eng-kit/.claude/skills/prompt-eval .claude/skills/prompt-eval
$ curl -fsSL --create-dirs -o .claude/agents/cost-watchdog.md https://raw.githubusercontent.com/Megha21-19/claude-code-data-eng-kit/main/.claude/agents/cost-watchdog.md $ curl -fsSL --create-dirs -o .claude/agents/eval-runner.md https://raw.githubusercontent.com/Megha21-19/claude-code-data-eng-kit/main/.claude/agents/eval-runner.md $ curl -fsSL --create-dirs -o .claude/agents/pipeline-reviewer.md https://raw.githubusercontent.com/Megha21-19/claude-code-data-eng-kit/main/.claude/agents/pipeline-reviewer.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"deny": [
"Bash(rm -rf:*)",
"Bash(databricks jobs run-now:*)",
"Bash(databricks workspace delete:*)",
"Bash(curl * | sh)",
"Bash(curl * | bash)"
]
},
"hooks": {
"PreToolUse": [
{
"matcher": "Write|Edit|MultiEdit|Bash",
"hooks": [
{
"type": "command",
"command": ".claude/hooks/pre-tool-protect-prod.sh"
}
]
}
]
}
} MCP servers (1)
| server | source | est. tokens |
|---|---|---|
| databricks-mini env: DATABRICKS_HOST, DATABRICKS_TOKEN | local / custom | 2.5k |
Skills (4)
Subagents (3)
| cost-watchdog model: claude-sonnet-4-5 | Estimates Databricks compute and LLM-API cost impact of a pipeline change. Spawn before any PR that adds a new job, changes cluster sizing, or modifies LLM-feature prompts. |
| eval-runner model: claude-sonnet-4-5 | Runs the prompt-eval harness for an LLM feature, compares against the baseline, and reports regressions. Spawn when a prompt or LLM-runner file changes. |
| pipeline-reviewer model: claude-sonnet-4-5 | Reviews PySpark pipeline changes for correctness, idempotency, and medallion-layer hygiene. Spawn before any PR that touches src/bronze, src/silver, or src/gold. |
Hooks (3)
| event | matcher | runs |
|---|---|---|
| SessionStart | * | .claude/hooks/session-start-context.sh |
| PreToolUse | Write|Edit|MultiEdit|Bash | .claude/hooks/pre-tool-protect-prod.sh |
| PostToolUse | Write|Edit|MultiEdit | .claude/hooks/post-edit-pyspark-checks.sh |
Slash commands (1)
/ship-checklist
Permissions
deny (5)
Bash(rm -rf:*)
Bash(databricks jobs run-now:*)
Bash(databricks workspace delete:*)
Bash(curl * | sh)
Bash(curl * | bash)
ask (0)
—
allow (15)
Read
Grep
Glob
Edit
Write
MultiEdit
Bash(git status)
Bash(git diff:*)
Bash(git log:*)
Bash(git branch:*)
Bash(ruff:*)
Bash(pyright:*)
Bash(pytest:*)
Bash(uv:*)
Bash(python:*)
Similar rigs
KimHance/claude-config-helper
Claude Code configuration helper plugin - review and generate CLAUDE.md, skills, agents, memory, hooks, settings, and MCP configs
Orchestrator 443 tok ·
rochaksharma19/claude-code-setup
Production-grade Claude Code configuration stack for AI/ML development
Pragmatist 28.1k tok ·
anomalyco/opencode
The open source coding agent.
Pragmatist 4.2k tok ·
DietrichGebert/ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
Pragmatist 1.6k tok ·