Data Engineer

Harness configuration for data engineers working with SQL pipelines, data lineage, and knowledge management. Includes lineage tracing, research tooling for documentation, and session capture for compounding institutional knowledge.

By harnessprotocol · 7 plugins bundled

Install

One-shot setup — download & apply the whole profile

Grab the profile's harness.yaml, then sync and compile. Works across Claude Code, Cursor, and Copilot at once.

$ npx @harness-kit/cli sync

$ npx @harness-kit/cli compile

compilewrites native config for Claude Code, Cursor, and Copilot automatically based on what's installed.

What's inside

Click any plugin to view its full documentation and security scan.

Workflow rules

Always trace lineage impact before modifying shared tables
Prefer CTEs over nested subqueries for readability
Document column-level transformations in commit messages

Knowledge seeding

Memory backend: memory-md

sql-patterns— Common SQL patterns and anti-patterns for data pipelines

data-quality— Data quality checks and validation strategies

View harness.yaml

version: "1"
kind: profile
metadata:
  name: data-engineer
  description: >-
    Harness configuration for data engineers working with SQL pipelines, data lineage, and knowledge management. Includes lineage tracing, research tooling for documentation, and session capture for compounding institutional knowledge.
  author:
    name: harnessprotocol
plugins:
  - name: lineage
    source: harnessprotocol/harness-kit
    version: "0.3.0"
  - name: research
    source: harnessprotocol/harness-kit
    version: "0.4.0"
  - name: orient
    source: harnessprotocol/harness-kit
    version: "0.2.0"
  - name: capture
    source: harnessprotocol/harness-kit
    version: "0.4.0"
  - name: explain
    source: harnessprotocol/harness-kit
    version: "0.2.0"
  - name: docgen
    source: harnessprotocol/harness-kit
    version: "0.2.0"
  - name: review
    source: harnessprotocol/harness-kit
    version: "0.3.0"
instructions:
  behavioral: |
    - Always trace lineage impact before modifying shared tables
    - Prefer CTEs over nested subqueries for readability
    - Document column-level transformations in commit messages