MapReduce root/orchestrator with a mandatory parallel Verify phase, narrative-first reduction, deterministic artifact lifecycle management (.rlm run/archives), and concurrency safety (per-run locks + cleanup lock). Use when coordinating many parallel map-worker tasks under optional hint_paths, then synthesizing narrative reports into a decision to iterate or finish.
[Extends backend-developer] Kafka specialist for Spring/Reactor. Use for Kafka producers/consumers, DLT, retry mechanisms, transactional outbox, event sourcing. Covers Spring Kafka 4.x and Reactor Kafka 1.3.x. Invoke alongside backend-developer.
Streaming, pagination, count strategies, bulk operations
Parallel thread/DuckLake discovery with XOR uniqueness from gay_seed. Finds "say" or MCP usage, cross-refs with all DuckDB sources, launches bounded parallel ops.
Create and maintain ExecPlans for complex work (design-to-implementation) following the repo's ExecPlan standard.
Validation rules, schema checks, duplicate detection, and coverage analysis for golden dataset integrity
Proactively analyzes Parquet file operations and suggests optimization improvements for compression, encoding, row group sizing, and statistics. Activates when users are reading or writing Parquet files or discussing Parquet performance.
Backup, restore, and validate golden datasets for AI/ML systems - ensuring test data integrity and preventing catastrophic data loss
Extract hidden Amp API patterns from local thread data via DuckDB analysis
FileSystemWatcher over /tmp with DuckDB/DuckLake persistence. Auto-starts on Amp sessions for resilient file monitoring with temporal queries.
Analyzes and assists with Fujitsu mainframe systems including FACOM, PRIMERGY, BS2000/OSD, OSIV/MSP, OSIV/XSP, NetCOBOL, PowerCOBOL, and Fujitsu JCL. Extracts business logic from Fujitsu COBOL programs, analyzes Fujitsu JCL jobs, migrates Fujitsu mainframe applications to modern platforms (Java, cloud, containers), and creates migration strategies. Use when working with Fujitsu mainframe migration, FACOM systems, BS2000, OSIV platforms, NetCOBOL, PowerCOBOL, Fujitsu-specific COBOL extensions, Fujitsu JCL, or when users mention Fujitsu mainframe modernization, analyzing Fujitsu COBOL/JCL, SYMFOWARE database, or planning migration from Fujitsu legacy systems.
Automatically triggered for database migrations, schema changes, and data transformations. Use when working with database structure, migrations, or ORM models.
Passthrough filter agent. Calls data sub-agents, validates responses, returns only clean data.
Compact context at natural breakpoints to free tokens and continue working. Use PROACTIVELY at phase boundaries, after commits, or when token usage >150k. Better than random auto-compact. Supports custom compaction prompts.
DuckDB time-travel queries for temporal versioning and causality tracking
Expert-level Looker BI, LookML, explores, dimensions, measures, dashboards, and data modeling
Aggregate and summarize event datasets (logs) using OPAL statsby. Use when you need to count, sum, or calculate statistics across log events. Covers make_col for derived columns, statsby for aggregation, group_by for grouping, aggregation functions (count, sum, avg, percentile), and topk for top N results. Returns single summary row per group across entire time range. For time-series trends, see time-series-analysis skill.
Generate interactive validation reports with quality scoring, missing data analysis, and type checking. Combines Pandas validation, Plotly visualization, and YAML configuration for comprehensive data quality reporting.
This skill should be used when the user asks to "work with polars", "create a dataframe", "use lazy evaluation", "migrate from pandas", "optimize data pipelines", "read parquet files", "group by operations", or needs guidance on Polars DataFrame operations, expression API, performance optimization, or data transformation workflows.
Transform data sources to Nixtla schema (unique_id, ds, y) with column inference. Use when preparing data for forecasting. Trigger with 'map to Nixtla schema' or 'transform data'.
Use this skill for processing and analyzing large tabular datasets (billions of rows) that exceed available RAM. Vaex excels at out-of-core DataFrame operations, lazy evaluation, fast aggregations, efficient visualization of big data, and machine learning on large datasets. Apply when users need to work with large CSV/HDF5/Arrow/Parquet files, perform fast statistics on massive datasets, create visualizations of big data, or build ML pipelines that do not fit in memory.
Distributed computing for larger-than-RAM pandas/NumPy workflows. Use when you need to scale existing pandas/NumPy code beyond memory or across clusters. Best for parallel file processing, distributed ML, integration with existing pandas code. For out-of-core analytics on single machine use vaex; for in-memory speed use polars.
