judgectl Agent Skill

Automating federal court data ingestion safely requires using judgectl to execute bounded runs with mutation safety controls like dry-run semantics or explicit --yes flags. This enforces fail-closed rules that block auto-publishing unverified claims.

Source validation for public legal sources is best handled through the judgectl interface, which performs source listing, validation, and info checks. It enforces evidence-first operational guardrails by treating evidence as authoritative and keeping memory derivative.

Yes, judgectl requires stable machine-readable JSON output for command-line operations. You must append --json to health checks and other commands to ensure the agent receives structured responses for source discovery and audit guardrails.

You can enable or disable data sources automatically by using explicit mutation controls. The judgectl interface requires explicit --yes flags for source mutations to prevent accidental or unauthorized data changes during the ingestion workflow.

The ingestion pipeline blocks auto-publishing unverified claims to enforce evidence-authoritative, fail-closed rules. This prevents accidental or unauthorized data mutations by ensuring the agent avoids inventing or asserting claims without verified snapshots.

Setting up audit guardrails for legal data review involves executing judgectl commands for health checks and source validation. The pipeline enforces evidence-first guardrails, ensuring memory remains derivative and treats verified snapshots as authoritative.