ghostinthedata-info

ghostinthedata-info/skills

Plugin Claude Code

Plugin marketplace listing 1 plugin: ghost-skills.

ghostinthedata-info/skills

Plugin Claude Code

ghost-skills.

ghostinthedata-info/skills

Instructions file

Instructions for ghostinthedata-info/skills, covering governance, agent skills, issue tracker, triage labels and domain docs.

ghostinthedata-info/skills

Skill Claude CodeCodex

Base orchestrator for designing analytics data models. Drives the Kimball grain-first workflow, profiling source data first, then routing to the right deep-dive — SCD1, SCD2, backfilling, and fact-table design. Use when the user is designing or refactoring dimensions, facts, a star schema, or a gold/presentation…

ghostinthedata-info/skills

Skill Claude CodeCodex

Classify data by sensitivity and apply the right controls — access control, masking/obfuscation, encryption, and auditing — driven by regulatory and confidentiality requirements. Covers the four A's (authentication, authorization, access, auditing), confidentiality levels, PII/PHI, and least privilege. Use when…

ghostinthedata-info/skills

Skill Claude CodeCodex

Elicit and pin down requirements for a data pipeline or model before any code. Nails grain, sources, consumers, definitions, freshness/volume SLAs, historization, access patterns, and acceptance criteria. Use when the user starts a new pipeline/model/report, or when scope is fuzzy and you need to prevent rework and…

ghostinthedata-info/skills

Skill Claude CodeCodex

Use this skill whenever a data or reporting incident needs to be identified, classified, triaged, communicated, resolved, or reviewed. Triggers include: a pipeline failure, delayed/late data, a transformation defect, a replication or ingestion failure, duplicate or missing records, a dashboard showing wrong or stale…

ghostinthedata-info/skills

Skill Claude CodeCodex

A methodology for tuning slow pipelines and queries — measure first, find the critical path, then fix with partition pruning, incremental processing, idempotent merges, and dependency cleanup. Tool-agnostic spine plus engine-specific reference files for traditional RDBMS (Oracle/SQL Server), cloud warehouses…

ghostinthedata-info/skills

Skill Claude CodeCodex

Elicit and pin down requirements for a data pipeline, model, or report before any code, then produce a solution-design doc. Nails value (pains/gains/profiles), grain, sources & authority, definitions, freshness/volume SLAs, historization, access patterns, and acceptance criteria. Produces a one-page-style design with…

ghostinthedata-info/skills

Skill Claude CodeCodex

Profile a new or unfamiliar dataset BEFORE building on it. Produces row counts, cardinality, null/blank analysis, key-uniqueness checks, distribution/outlier checks, type/precision audit, and an optional saved baseline for drift detection. Use when the user is about to model, ingest, or join a dataset they don't yet…

ghostinthedata-info/skills

Skill Claude CodeCodex

Build a defensible data test plan — what to test (uniqueness, referential integrity, nulls, accepted values, freshness, volume/variance), severity (error vs warn), and where tests run (Write-Audit-Publish). Use when adding tests to a model/pipeline, designing data-quality gates, or deciding what "good data" means here.

ghostinthedata-info/skills

Skill Claude CodeCodex

Turn analysis into a decision — work in the explanatory (not exploratory) space, lock Who/What/How, lead with the Big Idea, choose the right visual, declutter, and direct attention so the audience acts. Use when presenting results, building a dashboard, writing an analytical summary, or whenever a chart needs to drive…

ghostinthedata-info/skills

Skill Claude CodeCodex

Think in sets, not loops — replace cursors and row-by-row procedural code with declarative constraints, window functions, GROUPING SETS/CUBE/ROLLUP, CASE, set operations, and auxiliary tables. Use when SQL is being written procedurally, a transform loops over rows, or a query could be expressed declaratively.

ghostinthedata-info/skills

Skill Claude CodeCodex

Recognise and avoid the common SQL schema and query design traps — comma-separated lists, adjacency-only trees, missing foreign keys, EAV, polymorphic associations, FLOAT money, SELECT , unplanned indexes, and broken NULL logic. Use when designing a schema, reviewing a migration, or diagnosing a query that's slow…

ghostinthedata-info/skills

Skill Claude CodeCodex

Write portable, readable SQL — ISO-standard syntax, consistent ISO-11179 naming, set-oriented formatting, and constraints in the schema. Use when establishing SQL conventions for a repo, reviewing query craft, or naming tables and columns.

ghostinthedata-info/skills

Skill Claude CodeCodex

Structured, honest, empathetic feedback for data engineers at any level (junior, mid, senior, principal/staff). Identifies the engineer's expected maturity level from their position, then scores observed performance against SFIA-anchored, industry-benchmarked behaviours across technical, delivery/ops, collaboration…

ghostinthedata-info/skills

Skill Claude CodeCodex

Ultra-compressed communication mode. Cuts token usage 75% by dropping filler, articles, and pleasantries while keeping full technical accuracy. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman.

ghostinthedata-info/skills

Skill Claude CodeCodex

Teach the user a new skill or concept, within this workspace.

ghostinthedata-info/skills

Skill Claude CodeCodex

Create new agent skills with proper structure, progressive disclosure, and bundled resources. Use when user wants to create, write, or build a new skill.

ghostinthedata-info/skills

Skill Claude CodeCodex

Sets up an ## Agent skills block in AGENTS.md/CLAUDE.md and docs/agents/ so the ghost-skills pack knows this repo's warehouse platform & SQL dialect, transform/orchestration tooling, issue tracker (GitHub or local markdown), triage label vocabulary and domain-doc layout. Run before first use of profile-data…