Agent Control Plane — wireframe (wf4)

wf4 — net-new (Uber-cap idea): per-axis spend ceilings — per user / per agent / per tool — with cap coverage, who is over/near, and an "uncapped" worklist naming the entities with no ceiling (data-pipeline-agent = the LLM10 unbounded finding). Runs the same Audit→Assist→Enforce ladder; activates the previously-dead "Set ceilings" card.

wf4 — net-new: the four-question map (Assurance / Security / Identity & Access / Admin) over the five seats, linking every surface. The conceptual map the seat-only nav lacked.

wf4 — net-new: the 5th question, "is it working?" Task success / intervention / rework per agent & workflow, a fix-first roadmap (failed-task volume), and per-agent measured/inferred/none signal provenance. Performance becomes the speed/cost slice beneath it.

wf4 — adds a "wall mode" toggle: an ambient calm-vs-alarm status (one verdict + the few alarms that matter) to leave up on a screen, alongside the seat-lensed analyst view.

wf4 — adds lifecycle states (active / stale / forgotten / decommission / shadow) as a filter + per-row badge, surfacing the "old & forgotten" and orphaned agents to clean up.

wf4 — net-new: Teams & Users accountability roster, rolling agents/cost/at-risk-spend/findings up by owner (team rollup projected via directory). Distinct from Showback (per-identity spend).

wf4 — promotes the ghosted export to a concrete, previewable evidence pack: masthead, scope line, per-framework control evidence with why-trails, and an attestation footer — the document you would hand a regulator.

wf4 — carried forward from wf3 (seat overview), now under the four-question nav grouping with the Teams & Users surface available.

wf4 — carried forward from wf3 (seat overview), now under the four-question nav grouping with the Teams & Users surface available.

wf4 — carried forward from wf3 (seat overview), now under the four-question nav grouping.

wf4 — carried forward from wf3 (seat overview), now under the four-question nav grouping.

wf4 — carried forward from wf3 (seat overview), now under the four-question nav grouping.

wf4 — carried forward from wf3 (single-agent deep dive).

wf4 — carried forward from wf3 (governance-tagged spend, est honesty), now under the four-question nav grouping.

wf4 — carried forward from wf3 (risk-weighted per-identity spend); now paired with the net-new Teams & Users owner rollup.

wf4 — carried forward from wf3 (slow-AND-expensive, task-outcome gap noted).

wf4 — carried forward from wf3 (ungoverned-first worklist, framework grouping).

wf4 — carried forward from wf3 (Hardening-ROI fix-first, worst-case shadow pricing, trend).

wf4 — carried forward from wf3 (which/how-much, not contents; read/write asymmetry).

wf4 — carried forward from wf3 (per-agent dry-run, cost of flipping to Enforce).

wf4 — carried forward from wf3 (credential standard conformance, rotation age, compound badge).

wf4 — carried forward from wf3 (coverage %, shadow-agent export-gap warnings).

wf4 — carried forward from wf3 (governance-vs-spend story, risk concentration, backlog trend).