dekit

Build and maintain an internal technical LLM wiki when the user requests wiki work or a verified major or architectural change qualifies for automatic wiki review. Use for project memory, architecture notes, contracts, runbooks, decisions, glossary, incremental ingest, delta status, and wiki health. Do not use merely because a routine task or plan completed.

Explore ambiguous ideas and solution directions before evidence-backed research or implementation planning. Use for ideation, framing, assumption checks, option generation, product or architecture trade-offs, and deciding whether research or a plan is needed.

Research technical options with source-backed recommendations. Use for technology evaluation, architecture trade-offs, library/framework choices, best practices, scalability, security, maintainability, or when current external facts matter.

Create implementation plans, architecture decisions, and phased roadmaps. Use for Standard or Complex changes, new pipelines, schema changes, migrations, multi-file features, technology choices, or when acceptance criteria and verification strategy are unclear.

Review code and data pipeline changes for correctness, security, performance, maintainability, contract breaks, and missing verification. Use before merge, after implementation, for PRs, commits, pending diffs, or codebase risk scans.

Design data ingestion into landing/bronze layers. Use for source onboarding, transfer pattern selection, full/incremental/CDC scope, change detection, landing zones, schema drift, idempotent file or batch processing, and ingestion quality gates.

Design warehouse and semantic data models. Use for grain definition, fact/dimension design, star schema, Kimball, Data Vault, SCD, bridge tables, surrogate keys, semantic metrics, and model review.

Define and enforce data quality rules, contracts, assertions, reconciliation, quarantine, and quality gates. Use for completeness, uniqueness, validity, freshness, schema, referential integrity, SCD checks, or source-to-target validation.

Design, implement, and review DataOps workflows for data platforms. Use for CI/CD for data pipelines, infrastructure as code, deployment automation, monitoring, alerting, rollback plans, cost controls, secrets management, and platform reliability.

Debug with root-cause analysis before fixes. Use for bugs, failing tests, CI failures, runtime errors, Spark/SQL performance, schema drift, data quality incidents, logs, metrics, and unexplained behavior.

Search library/framework documentation via llms.txt (context7.com). Use for API docs, GitHub repository analysis, technical documentation lookup, latest library features.

Create and maintain end-user or public-facing documentation. Use for user guides, product docs, API consumer docs, onboarding docs, tutorials, release notes for users, and documentation meant to be read outside the engineering team.

Git operations with conventional commits. Use for staging, committing, pushing, PRs, and merges.

Write and organize production-grade notebooks across Fabric, Databricks, and Jupyter. Use for notebook cell structure, parameters, idempotency, platform utilities, validation cells, notebook-to-production conversion, and notebook hygiene.

Fast codebase scouting for file discovery, task context gathering, and quick searches across directories.

Run security review for code, data pipelines, infrastructure, notebooks, and configs. Use for STRIDE/OWASP checks, secrets, auth/authz, PII exposure, injection, supply chain, IAM, encryption, audit logging, and security remediation planning.

Write and optimize PySpark and Spark SQL. Use for Spark DataFrames, joins, windows, UDF decisions, partitioning, caching, AQE, explain plans, OOM, shuffle, skew, Delta writes, and Spark pipeline performance.

Write and review complex SQL across dialects. Use for CTEs, windows, pivots, temporal joins, recursive queries, MERGE/upsert, SCD logic, dialect translation, query optimization, and SQL anti-pattern detection.

Run and design verification for code, data pipelines, SQL, Spark, notebooks, and UI changes. Use for unit, integration, e2e, schema, row count, reconciliation, idempotency, coverage, build, and QA reports.