Agent skill · software engineering · jeremylongshore
juicebox-ci-integration
Configure Juicebox CI/CD.
Why this skill is useful
Adds a CI/CD configuration and integration testing scripts specifically for Juicebox, which the AI wouldn't generate reliably on its own.
What it needs
Requires npm installed locally. Requires juicebox account access. About 2k tokens when loaded. Last updated 2026-08-07. 2,604 stars on the source repository.
What this skill does
Juicebox CI Integration Overview Set up CI/CD for Juicebox AI data analysis integrations: run unit tests with mocked dataset and analysis responses on every PR, validate live API connectivity for data queries on merge to main. Juicebox provides AI-powered data exploration and visualization, so CI pipelines verify dataset upload logic, analysis execution, and result parsing workflows. GitHub Actions Workflow Mock-Based Unit Tests Integration Tests Error Handling CI Issue Cause Fix ---------- ------- ----- 401 Unauthorized Invalid API key Regenerate at juicebox.ai account settings Analysis stuck on processing Large dataset or complex query Increase polling timeout to 120s Dataset not found (404) Dataset ID changed or deleted Use listDatasets to get a valid ID dynamically Rate limit (429) Too many concurrent analyses Queue analyses and limit to 2 parallel runs Empty insights array Insufficient data for AI analysis Ensure test dataset has 100+ rows with varied data Prerequisites CI secret references, a sandbox workspace, synthetic prospects, protected branches, source-authority checks, and a rollback mechanism. Instructions 1. Run mocked tests first, including unknown source, denied scope, suppression, malformed enrichment, quota, and export-block cases. 2. Run a bounded sandbox integration with idempotency; prohibit production sources/destinations and literal credentials in CI. 3. Emit aggregate counts, opaque IDs, and policy revisions only; fail for unapproved source, expanded scope, contact-data leakage, or export path. 4. Canary after protected review, verify contactsexported=0, and restore the last-known-good revision on failure. Output Publish a CI receipt with commit SHA, fixture revision, sandbox workspace, test totals, source/destination/suppression checks, export-count assertion, canary outcome, and rollback reference. Exclude contacts, enrichment fields, and secrets. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills jeremylongshore/juicebox-ci-integration--1cf2ca