Ask Jev

Use on an explicit Jev request, including /ask-jev <question> or Pi's /skill:ask-jev <question>, or a bounded policy-authorized assessment handoff. A Pi-expanded skill block counts as explicit invocation only when it originates from the user's command; generated command text is not consent. Follow-ups retain the original scope and spent allowance. Resolve a missing question before proceeding.

Before credential access or authenticated operations, read authorization and shared hosted limits. Loading the skill, related conversation and configured credentials grant no permission. A denied autonomous route returns no-call; ordinary engineering continues where sufficient, or reports a material evidence gap. Authorized use needs no repeated confirmation inside valid bounds.

The agent gathers evidence and explains the result; Jev judges only the state sent to TypeSafe's hosted service. It cannot independently inspect the repository, conversation, filesystem, or telemetry. Explicit invocation covers the named normal billable assessment under package bounds and applicable project restrictions, not installations, expanded sensitive disclosure or consequential actions.

Project context

For project work, reuse applicable project instructions and the existing working agreement once per context. Check relevant decision/effect authority, safety, verification/review, documentation and continuation guidance; equivalent layouts are valid. Missing or conflicting common guidance blocks dependent project work: name the gap and recommend explicit setup/reconciliation without starting it or inventing policy. A conditional tracker, model or domain gap blocks only that capability. A skill update does not change project permissions.

A named standalone assessment uses request-specific scope without manufacturing a project agreement. Neither setup nor ordinary skill invocation opts into an assessment. The authorization reference still governs exact invocation/policy, disclosure, shared reservations and lifecycle; no delivery entry is needed to obtain model policy. Missing autonomous permission remains no-call, and ordinary work needs no model result.

Workflow

1. Route the original question

Use Jev for bounded semantic judgments with an explicit standard: classification, routing, policy or requirement checks, evidence support, or described rubric levels.