keboola-cli_skill

Overview

This skill provides practical guidance for working with Keboola project configurations, local .keboola repositories, and component config.json files. It helps you inspect, edit, validate, and troubleshoot transformations, extractors, writers, and orchestrations. Use it to speed up safe edits and maintain consistent pipelines across branches.

How this skill works

It inspects the local project structure under .keboola, reads manifest.json, and opens per-branch component directories to parse config.json and meta.json. It explains common component schemas (transformations, extractors, writers, orchestrations), storage mappings, and the typical locations for scripts, blocks, and rows. It also recommends validation and diff commands to run before pushing changes.

When to use it

• When adding or modifying transformation SQL/Python/R blocks in keboola.snowflake-transformation
• When mapping input/output tables for extractors or writers in config.json
• When reviewing branch-specific configuration under .keboola/[branch]/[component]/[config]/
• When troubleshooting invalid JSON, push conflicts, missing tables, or mapping issues
• When creating or updating orchestrator tasks and schedules

Best practices

• Always run kbc diff before and after edits and kbc validate before pushing
• Keep meta.json name and description up to date for discoverability
• Break complex transformations into multiple blocks with meaningful names
• Use consistent naming conventions for storage input/output destinations
• Validate JSON syntax and test transformation logic locally where possible

Example use cases

• Add a new input table to a transformation by updating storage.input.tables with source and destination
• Add an output table and primary key in storage.output.tables for downstream writers
• Edit a Snowflake transformation script by modifying the script array inside a block/code section
• Inspect .keboola/manifest.json to confirm project ID, API host, and branch info before pushing
• Resolve push conflicts by pulling latest branch, merging config.json changes, and validating