# PM View (/docs/views/pm)



The **PM** runs a Pod: the node above the Books. It never authors a ticket. Its actions are
*allocations* — sizing children's envelopes, arming and de-arming, assigning Coverage, narrowing
authority downward. The PM's job splits cleanly into two halves at two timescales, and each half
reads its own screen.

## The two halves of the PM [#the-two-halves-of-the-pm]

* **Discovery — the Scanner.** When a universe-wide scan fires, the PM's discovery half does the
  single-name deep read that decides whether to author a new leaf into the pod. That is the
  [Scanner View](/docs/views/scanner), and it ships today: a scan-fire wake seated to the PM
  resolves to it.
* **Allocation — the aggregates cockpit.** Between discoveries, the PM reads the pod's aggregate
  exposure and envelope utilization across its children and moves budget. That is the cockpit this
  page describes.

The two halves are one seat because they share a role — the PM that allocates is the PM that
discovers — but they are different jobs at different speeds, which is why they read different
screens. The discovery half is a minutes-latency single-name read; the allocation half is a slow,
aggregate read.

## The aggregates cockpit [#the-aggregates-cockpit]

The allocation cockpit is the `PodView` — the screen that renders a Pod node's aggregate state.
Two invariants are fixed by contract and shape the cockpit:

* **A `PodView` has no kernel, by construction.** It allocates and aggregates; it never holds
  positions. There is no acting block on a Pod node, because a Pod node acts by moving envelopes,
  not by resting orders.
* **Positions appear only at Books.** The leaf is the only place inventory lives; every node above
  a Book allocates and aggregates. A fact no child has published is **absent** — and absent is
  UNKNOWN, which de-arms a PM wake with a logged reason, never a silent zero. The cockpit reads its
  children's real facts or it reads an honest gap; it never fills one in.

The cockpit rides the same View grammar and the same resolution rules as the seat Views: it is a
role-keyed screen, chosen by the same `explicit View > seat View > phase default` precedence, and
enabled by the same `seatViews` opt-in. As the aggregate panes land, the cockpit joins this
section as the PM's allocation View, alongside the Scanner View that already covers its discovery
half.

## Open vs wake [#open-vs-wake]

The PM allocates on the pod's clock, not the tape's — it wakes to re-size an envelope, to de-arm a
child whose triggering fact went absent, or to review aggregate exposure, rather than to manage a
tick. Where the [Watcher View](/docs/views/watcher) leads with what moved on a single name, the
allocator's read leads with exposure and utilization summed across children — the aggregate the
Books roll up to. That aggregate is the **variant-independent** layer of the section: a Pod
allocates the same way whether its Books trade equity-index options, futures, perpetuals, or FX,
because it reads envelopes and exposure, not chains. The instrument variant lives at the Book; the
allocation cockpit sits above it.