Bound managed Jev requests and align the individual usage contract

[Coding-Dev-Tools]

Summary

Bounds the client managed Jev transport and keeps managed requests within the supported operation schemas and request limits. The integration preserves deterministic/local fallback, caller consent, tenant/session checks, and per-user accounting. It does not expose a general-purpose provider proxy.

The stack documents the same individual managed allowance for every eligible Pro user and Team named seat, including paid viewers with active paid, trial, or test entitlements. Managed Jev remains disabled pending service acceptance.

Managed Jev contract

When available, each eligible individual receives 100 evaluated questions per rolling hour, 1,000 per rolling five hours, and 2,000 per rolling 24 hours. All three independent limits apply per person; usage is not pooled or monthly. Each evaluated question counts once, batches cannot bypass the limits, and command review uses two questions. Customers pay no extra charge and do not supply a Jev provider key.

The existing production fleet guard remains 100 questions per day and conflicts with the individual limits. Cloud PR #102 is currently at bc254d64, which pools Team usage and omits the five-hour limit, so managed access remains unavailable pending a compliant server change and capacity qualification. No accuracy or cost-saving benefit is claimed.

Verification

• Head: 56625f4b98b1613e263ed14c4190f81a893dd3f7.
• The new planner-fallback and compact-preview regressions plus documentation/link/checksum checks passed in the local direct harness; Ruff passed on the changed Python files.
• CI and CodeQL are running on this exact head.
• No real provider calls, service activation, deployment, or billing changes are included.