fix(pane): show a task's next move instead of dead text or a fixed method

.agents/skills/fleet-pane/SKILL.md

@@ -89,10 +89,14 @@ became one. Every word on it comes off `task.yaml`'s `publish` block, written by
 pane calls no `gh` and says nothing `queue.sh show` would not print in the same
 word. Colour carries the verdict, and **`green` is not the ok colour**: it means
 every gate the forge knows about holds and nobody vetted it, which is a
-different claim from `ready` and is why fleet will not merge it for you. A
-narrow column drops the parts in a fixed order — the `— yours to merge` note,
-then the method, then the age, then the `#44` — and the link survives losing its
-label.
+different claim from `ready` and is why fleet will not merge it for you. The
+note beside a state is its next move, not its colour — `— yours to merge` on
+green, `— review` on `open`, since `open` is a fact `collect` proved, not a
+verdict, and gets no colour that would claim one. A narrow column drops the
+parts in a fixed order — the method first, then the note, then the age, then
+the `#44` — and the link survives losing its label. `interface/fleet_queue.lua`
+(`PUBLISH_WORD`, `PUBLISH_LADDER`) is the owner of that order and argues it in
+place.
 
 **No row is drawn for a task with nothing to report about its publish**: a
 record from before `publish` existed, or a task whose publish has not started.

interface/fleet_queue.lua

@@ -126,6 +126,14 @@
 -- A task with an artifact is one row SHORTER than it was before this row
 -- existed, and no task is taller.
 --
+-- AND IT SAYS THE NEXT MOVE, NOT ONLY THE STATE. `open` and `green` are facts
+-- about a pull request; `review` and `yours to merge` are what the operator
+-- does about them, and colour can carry the first but never the second. That
+-- word used to be the first thing the row's width ladder dropped, so it was
+-- drawn at neither width `scripts/pane-selftest.sh` renders while the publish
+-- METHOD — one unchanging word under every task in the queue — was drawn at
+-- both. The ladder gives up provenance before it gives up the action now.
+--
 -- FOLLOW-UP, WRITTEN DOWN RATHER THAN DONE: the probe below should become
 -- `queue.sh list --tsv`. That would make "this pane cannot disagree with
 -- `list`" literal instead of argued, and it would drop a dozen `sed`/`awk`/
@@ -873,20 +881,36 @@ local BLOCKER_KIND = {
 --- here, or by repainting `green` with `theme.ok` because warn "looks like a
 --- problem" — converts fleet's evidence-over-trust property back into trust,
 --- silently, and it is the sharpest risk this row's design names. The note
---- beside it says whose job the merge is, and it is the first thing the ladder
---- below drops.
+--- beside it says whose job the merge is, which the ladder below now keeps
+--- ahead of the method rather than dropping first.
 ---
 --- `open` is muted rather than ok on purpose: `collect` proved the pull request
 --- exists and comes from this task's branch, and nothing has yet looked at its
 --- checks. That is a fact, not a verdict, so it gets no colour that reads as one.
+--- Muted next to a collapsed `n landed` row reads as SETTLED, though, and it is
+--- the opposite: `open` is the state every task passes through on its way to a
+--- review nobody has done yet. So it carries a NOTE instead of a colour — the
+--- one mechanism here that can say what to do without claiming a verdict, and
+--- `green` already used it. The note is what `descriptors` says of a `done`
+--- task in its own words: the pull request is open and the next move is to
+--- review it. One word, because the ladder below keeps a note only while the
+--- whole row fits, and thirty columns is the width this pane is usually given:
+--- `open — review` clears it and `open — review it` misses it by one column.
+---
+--- `green` KEEPS ITS LONGER NOTE AND THEREFORE LOSES IT AT THIRTY. `yours to
+--- merge` will not fit beside a reference and an age in a 30-column column, and
+--- the obvious shortening — `merge` — puts a word one letter from `merged` in
+--- the same column that draws `merged`, which is a worse row than no note. It
+--- says whose job the merge is, which is the half `green` itself cannot carry,
+--- and that half is worth more at 44 than a hazard is at 30.
 ---
 --- A state this table does not know is drawn verbatim and muted — the same rule
 --- `BLOCKER_KIND` follows, and for the same reason: an unknown word is shown,
 --- never mapped to a plausible one.
 local PUBLISH_WORD = {
   unverified = { text = "UNVERIFIED", tone = "bad" },
   unknown = { text = "unknown", tone = "muted" },
-  open = { text = "open", tone = "muted" },
+  open = { text = "open", tone = "muted", note = "review" },
   pushed = { text = "pushed ✓", tone = "ok" },
   draft = { text = "draft", tone = "warn" },
   ["checks-running"] = { text = "checks ⟳", tone = "warn" },
@@ -907,17 +931,26 @@ local PUBLISH_GLYPH = "⇡"
 --- What the publish row gives up as the column narrows, in order.
 ---
 --- The same shape `note_spans` uses — drop in a fixed order, truncate last —
---- and the order is what each part is FOR. The NOTE goes first: it is a
---- sentence about whose job a merge is, and the coloured word already carries
---- the fact. The METHOD next, because it is a property of the task that never
---- changes and the pull request page says it anyway, while the STATE is the
---- part an operator acts on. Then the AGE, then the artifact REFERENCE — which
---- is a label for the link this row carries, and the link survives losing its
---- label. Then the glyph. The state word is the last thing standing, and it is
---- truncated only when the column is narrower than the word itself.
+--- and the order is what each part is FOR. The METHOD goes first, because it is
+--- a property of the task that never changes, the brief that dispatched the task
+--- declared it, and the pull request page says it anyway. The NOTE next. Then
+--- the AGE, then the artifact REFERENCE — which is a label for the link this row
+--- carries, and the link survives losing its label. Then the glyph. The state
+--- word is the last thing standing, and it is truncated only when the column is
+--- narrower than the word itself.
+---
+--- THE NOTE USED TO GO FIRST AND THEREFORE NEVER WENT ANYWHERE. `yours to
+--- merge` costs 21 columns beside its word, which put the full rung over budget
+--- at 44 and far over it at 30 — the two widths `scripts/pane-selftest.sh`
+--- renders, and 30 is what this pane routinely gets. So the one segment that
+--- answers "what do I do about this" was drawn at no width an operator has,
+--- while `no-mistakes` — the same word under every task in a fleet with one
+--- publish method — was drawn at both. The note is the row's ACTION and the
+--- method is its provenance; provenance is what a column this narrow can afford
+--- to lose.
 local PUBLISH_LADDER = {
   { note = true, method = true, ref = true, age = true, glyph = true },
-  { method = true, ref = true, age = true, glyph = true },
+  { note = true, ref = true, age = true, glyph = true },
   { ref = true, age = true, glyph = true },
   { ref = true, glyph = true },
   { glyph = true },
@@ -1192,6 +1225,13 @@ end
 --- something about, and muted for a fact with no verdict attached. The table
 --- above owns which is which, and owns the argument for `green`.
 ---
+--- AND THE NOTE CARRIES THE NEXT MOVE, WHICH COLOUR CANNOT. A tone says how
+--- worried to be; it cannot say `review` rather than `merge`, and those are
+--- different actions on rows that sit two lines apart. It is also the only way
+--- to say anything about a state whose whole point is that it has no verdict:
+--- `open` stays muted and still says what to do. The ladder above keeps it
+--- ahead of the method for that reason.
+---
 --- THE AGE IS THE AGE OF THE LOOK, not of the task. The shepherd runs on the
 --- lead's cadence rather than on a clock, so a `checks ⟳` recorded forty
 --- minutes ago has to read forty minutes old — a state word with no age

scripts/lib/pane_harness.lua

@@ -219,6 +219,12 @@ _G.thurbox = { taken_at_ms = NOW * 1000, sessions = {}, runs = {} }
 -- a RUNNING topic that also holds a task the forge already merged, a chain of
 -- blockers of which one is long since cleared, tasks nothing has emitted an
 -- event for, and one settled topic underneath.
+--
+-- It also holds the publish state EVERY task passes through and no fixture used
+-- to reach: `open`, on a task `shepherd` linked by head branch before `collect`
+-- read the worker's result. That is why the pane's most common publish row went
+-- unrendered by anything that could be argued with, and it is the row the
+-- operator's next move — review it — hangs off.
 local function ago(minutes)
   return NOW - minutes * 60
 end
@@ -275,7 +281,9 @@ local TOPICS = {
       },
       {
         id = "03-drop-the-webui-selftest", state = "dispatched", title = "Drop the monitor's selftest and its CI job",
+        artifact = "https://github.com/Thurbeen/fleet/pull/52",
         brief = 1, events = 2, result = 1, moved = ago(3),
+        publish = { "no-mistakes", "open", ago(3) },
       },
     },
   },

scripts/pane-selftest.sh

@@ -24,6 +24,12 @@
 #   4. THE TWO TALLIES AGREE. The counter row counts TASKS by state; a section
 #      heading covers TOPICS. They were both drawn as bare numbers, and
 #      `4 running` above `RUNNING 5` is the pane contradicting itself.
+#   5. THE PUBLISH ROW SAYS THE NEXT MOVE, AT BOTH WIDTHS. A state word is a
+#      fact about the pull request; what the operator does about it is a
+#      second thing, and the row's `note` is where it lives. The note was the
+#      FIRST thing the width ladder dropped, so it reached neither width this
+#      file renders while `no-mistakes` — one word under every task in the
+#      queue — reached both. These assert the order that fixes it.
 #
 # And the rule none of that may cost: it still degrades. The pane routinely
 # gets thirty columns, so every assertion here runs at 30 as well as at 44.
@@ -117,6 +123,23 @@ expect "a result nothing has collected is still said" "uncollected" "$WIDE"
 expect "the pull request is still named" "#47" "$WIDE"
 expect "the publish verdict is still drawn" "green" "$WIDE"
 
+# --- 5. the publish row says the next move ----------------------------------
+
+# `open` is the state EVERY task passes through — `collect` proved the pull
+# request exists and nothing has looked at its checks yet — and it is drawn
+# muted, because a fact is not a verdict. Muted beside a collapsed `n landed`
+# row reads as settled, which is the opposite of what it means, so the row says
+# the move in words instead of borrowing a colour that would claim one.
+expect "an open pull request says what to do about it" "open — review" "$WIDE"
+expect "and the second pull request is named" "#52" "$WIDE"
+expect "a green one still says whose merge it is" "green — yours to merge" "$WIDE"
+
+# The order the ladder gives things up in, pinned: the METHOD is provenance and
+# the NOTE is the action, so a row too narrow for both keeps the action. Before
+# this, `no-mistakes · #47 · green` is what 44 columns drew and the note was
+# drawn at no width at all.
+refute "the method gives way to the note, not the reverse" "no-mistakes · #47" "$WIDE"
+
 # A blocker recorded against a task that is NOT waiting holds nothing —
 # `queue.py` clears a `queued` task and no other — so drawing it would explain
 # why a task is stuck about a task that is running.
@@ -162,5 +185,8 @@ fi
 expect "the running work is still named at 30" "Cut the pane back" "$NARROW"
 expect "the blocker that holds is still there at 30" "↳ 02-shepherd" "$NARROW"
 refute "and still no cleared blocker" "✓ 01-declare" "$NARROW"
+# The width the pane routinely gets, which is the whole reason the note moved
+# up the ladder: a next move drawn only at 44 is a next move nobody reads.
+expect "the next move survives 30 columns" "open — review" "$NARROW"
 
 exit "$failed"