Agent skill · operations · jeremylongshore
juicebox-incident-runbook
Juicebox incident response.
Why this skill is useful
Provides specific diagnostic commands and incident response procedures for the Juicebox API that the AI wouldn't reliably generate on its own.
What it needs
Requires jq installed locally. Requires juicebox account access. About 3k tokens when loaded. Last updated 2026-08-07. 2,604 stars on the source repository.
What this skill does
Juicebox Incident Runbook Overview Incident response procedures for Juicebox AI analysis platform integration failures. Covers analysis timeouts, dataset corruption, quota exhaustion, and export failures. Juicebox powers AI-driven people search and candidate analysis, so incidents disrupt recruiting pipelines, talent intelligence workflows, and automated sourcing. Classify severity immediately using the matrix below and follow the corresponding playbook. Severity Levels Level Definition Response Time Example ------- ----------- --------------- --------- P1 - Critical Full API outage or dataset corruption 15 min Health endpoint returns 5xx, analysis results missing P2 - High Analysis timeouts or export failures 30 min Search queries hang beyond 30s, CSV exports fail P3 - Medium Quota exhaustion or rate limiting 2 hours 429 responses, account quota at 100% usage P4 - Low Partial data or degraded result quality 8 hours Search returns fewer results than expected Diagnostic Steps Incident Playbooks API Outage 1. Confirm via health endpoint and status.juicebox.ai 2. Activate fallback mode — serve cached search results to active users 3. Pause any automated sourcing pipelines to avoid wasting quota on retries 4. Notify recruiting team that live search is temporarily unavailable 5. Monitor status page and resume operations once health check passes Authentication Failure 1. Verify API key is set: echo $JUICEBOXAPIKEY wc -c 2. Test with health endpoint (see diagnostics above) 3. If 401: API key may be revoked — regenerate in Juicebox dashboard 4. If 403: check account tier permissions for the requested endpoint 5. Deploy new key and verify search requests succeed Data Sync Failure 1. Check if recent analysis results are returning stale or incomplete data 2. Verify export endpoints are responding — test a small CSV export 3. If exports fail: check if the analysis job completed successfully first 4. For dataset corruption: re-trigger the analysis with fresh parameters 5. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills jeremylongshore/juicebox-incident-runbook