Agent Control Plane — persona-first synthesis (wf5)
wf5 — the "best of wf3 + wf4" synthesis: organized by PERSONA (wf3's clean per-seat flat nav, seat-lensed deep pages) while carrying wf4's richer surfaces (Effectiveness, Usage Limits, Teams & Users, lifecycle inventory, wall mode, evidence pack). The four questions are demoted to an IMPLICIT page ordering (healthy → working → safe → accountable → traced → wired) — no group headers, no Orientation page. Net-new: a behavioral Activity Log ("what happened?") — the flight recorder behind every finding — on the Security / Engineering / Compliance seats, plus effectiveness + activity woven onto the agent profile, the findings→activity bridge, and an "are they earning their keep?" ROI line on the Executive board.
Phase 2 — persona-partitioned agent-native wireframes anchored on the qcontrol control plane.
Design Specs
Pages
wf5 — net-new: the "what happened?" flight recorder. Reconstructed agent → action → target → outcome decision log (filter by family, flagged-only, ties-to-a-finding), event-volume timeline + family mix, per-agent activity rollup, and the honest "dark agents leave no trail" gap. The full record behind the distilled Findings worklist; on the Security / Engineering / Compliance seats, seat-lensed (incident-response / debug / audit).
wf5 — carried from wf4, now woven with the new questions: an "is it working?" effectiveness card (success / intervene / rework / quality with signal provenance) and a "what it did" recent-activity slice that replays into the Activity Log.
wf5 — carried from wf4 (seat-lensed worklist) plus a bridge to the new Activity Log: "this is the distilled worklist — replay the actions behind a finding in the full record."
wf5 — carried from wf4 plus an "are the agents earning their keep?" ROI line: fleet success rate, intervention rate, spend-per-successful-task, and measured-coverage — the board-level read of the "is it working?" question.
wf5 — carried from wf4: the "is it working?" surface (task success / intervention / rework per agent & workflow, fix-first by failed-task volume, measured/inferred/none provenance). Now a persona surface on the Engineering & Executive seats rather than a four-question group member.
wf5 — carried from wf4 (seat-lensed analyst view + ambient wall mode).
wf5 — carried from wf4 (seat-lensed roster + lifecycle states: active / stale / forgotten / decommission / shadow).
wf5 — carried from wf4 (per-axis spend ceilings, cap coverage, uncapped worklist, Audit→Assist→Enforce).
wf5 — carried from wf4 (accountability rollup: agents/cost/at-risk-spend/findings by owner).
wf5 — carried from wf4 (previewable evidence pack: masthead, scope line, per-framework why-trails, attestation).
wf5 — seat overview, persona-first nav (flat page list, Activity Log added to the seat).
wf5 — seat overview, persona-first nav (flat page list).
wf5 — seat overview, persona-first nav (flat page list).
wf5 — seat overview, persona-first nav (flat page list, Activity Log added to the seat).
wf5 — seat overview, persona-first nav (flat page list, Activity Log added to the seat).
wf5 — carried from wf4 (governance-tagged spend, est honesty).
wf5 — carried from wf4 (risk-weighted per-identity spend; paired with Teams & Users).
wf5 — carried from wf4 (slow-AND-expensive; sits beneath Effectiveness as the speed/cost slice).
wf5 — carried from wf4 (Hardening-ROI fix-first, worst-case shadow pricing, trend).
wf5 — carried from wf4 (which/how-much, not contents; read/write asymmetry).
wf5 — carried from wf4 (per-agent dry-run, cost of flipping to Enforce).
wf5 — carried from wf4 (credential standard conformance, rotation age, compound badge).
wf5 — carried from wf4 (coverage %, shadow-agent export-gap warnings).