generated: '2026-07-21'
method: searched
source: https://github.com/clickzetta/clickzetta-skills
description: >-
  ClickZetta publishes a first-party Agent Skills library (SKILL.md format,
  consumable by Claude Code, Codex, and the czcode agent). The repo turns
  ClickZetta operational knowledge into ~28 task-oriented skills plus reference
  docs. These are provider-published; this index references them verbatim at the
  source rather than duplicating each SKILL.md.
repo: https://github.com/clickzetta/clickzetta-skills
skills:
- name: clickzetta-file-import-pipeline
  summary: Import files from URLs, local paths, and Volumes into Lakehouse tables.
- name: clickzetta-kafka-ingest-pipeline
  summary: Kafka ingestion via READ_KAFKA Pipes.
- name: clickzetta-batch-sync-pipeline
  summary: Studio offline batch sync tasks.
- name: clickzetta-cdc-sync-pipeline
  summary: Multi-table CDC sync from MySQL / PostgreSQL.
- name: clickzetta-dbt-modeling
  summary: Source discovery, model design, and incremental materialization with dbt.
- name: clickzetta-dynamic-table
  summary: Create and configure Dynamic Table refresh; best practices.
- name: clickzetta-semantic-view
  summary: Define logical tables, dimensions, and metrics as Semantic Views.
- name: clickzetta-zettapark
  summary: ZettaPark DataFrame API — Session setup and transformations.
- name: clickzetta-spark-flink-connector
  summary: Spark / Flink read and write operations against Lakehouse.
- name: clickzetta-ai-function
  summary: Use AI_COMPLETE and AI_EMBEDDING SQL functions.
- name: clickzetta-volume-manager
  summary: External / User / Table Volumes and file operations.
- name: clickzetta-table-lineage
  summary: Lineage visualization via information_schema.job_history.
notes: >-
  Full list is 28 skills plus official docs and one independent skill; the
  entries above are representative marquee flows. See the repo for the complete,
  authoritative SKILL.md set.