The package landed here with one store spec; the rest of its tests were written
three layers up, so this layer and the two above it failed the per-file
coverage gate on 191 locations while every test passed.
Adds the node half's invariant companion, the two components, the registration
apply, and the stores' failure paths. `currentValue` drops a `?? ''` that an
empty roster already returned before reaching — the same shape the later layer
arrived at, so the two converge instead of conflicting.
Six review findings on the select surface, all reachable from the wire:
**Resume read the header, not the log.** The switch was recorded as
`agent-preset/selected` and every projection resolved from it, but `agentFor`
still composed from `inspected.meta.agentPreset` — the value written once at
creation. A blank session that switched and then ran turns came back after a
restart under the ORIGINAL preset, restoring that history under the tool set it
was not produced with, which is the mismatch this feature exists to prevent.
`inspected` already carries the events.
**Cold summaries dropped the preset entirely.** `summarizeCold` hand-copied
three header fields and omitted the fourth, so a restored session reported no
preset and the picker showed the deployment default. It now uses the same
projection the attached path does.
**`select` had no gate.** Two concurrent selects both passed the blank check;
the second `unmountPresetFor` then found no record, because the first had
already removed it, and both mounts installed into one agent layer. Selects on
one session now queue, and the blank check is re-read inside the queue. This is
not turn admission — a `session.prompt` racing a switch is the agent loop's to
reserve — but it closes the select-versus-select tear-down.
**A same-id restore was skipped.** The roster is a live directory, so "the same
inputs that worked a moment ago" does not hold: a changed file is exactly how a
same-id reselect fails, and skipping the restore left the agent with no
composition at all.
**`writable` was dead state**, initialized true and never set, so the row could
never disable. It now carries `settings.describe`'s bit — a browser that may
not write settings sees the current default and no control, rather than one
whose write answers `settings-not-exposed`.
**`list` was documented as id-ordered.** It is root-precedence order with each
root's own presets sorted, first root to supply an id winning.
`ui-question`'s node half called `ctx.tools.register` on the host context.
`ScopedLayers.merge()` combines the global layer with the agent's exact-scope
layer, and an unscoped registration lands in the global one — so the tool
reached every agent no matter which preset composed it. `core-web`, sold as a
two-tool benchmark surface, really presented three.
Rendering a question is a host UI capability; having the tool is an agent
capability, and only a preset decides that. The node half is now empty and the
`tool-ask-user` row moved into the preset that wants it. The TUI keeps its own
row, having no presets.
The composition tests now assert the global tool layer is EMPTY, which is the
invariant that would have caught this: any tool outside a preset reaches every
agent. The browser lane's composition, seeded-history, and hermetic-skill
assertions address their registries through a composed agent for the same
reason — those services are per session now, and the host cannot resolve an
`isolate` realm by name.
The creation header names the preset a session STARTED with and is frozen,
which is correct — it is a creation fact. Switching is legal only while a
session is blank, and that looked like enough: no history exists yet.
It is not, because the switch's effect outlives the blank window. The user
switches, then sends the first message; every turn from there runs under the
new composition while the header still names the old one. The session is
then locked around a misrecorded preset, and resume reads the header to
rebuild it — composing one preset's tools over a history another produced,
which is exactly the replay the blank-only lock exists to prevent, reached
by another route. A picker showed `standard` for a session running
`core-web`.
A switch is now an `agent-preset/selected` event appended after the swap
commits, and `resolveSessionPreset()` (last selection, else the header) is
what every reconstruction reads: the summary, resume, the conflict guard,
and the fork introduced one layer down.
A seat in the composer tool row, left of the model select, showing the preset
THIS session runs — its own recorded one, not the deployment default, because a
resumed session runs what it was created with. `SessionSummary` carries
`agentPreset` for that, alongside `cwd` and `origin`.
The switch exists only while the conversation has not started. After the first
turn the seat becomes a plain label rather than a disabled menu: a greyed
control reads as "temporarily unavailable", when in fact the choice is gone for
good. The seat never asks in that state, and the host refuses independently
with `agent-preset-locked`, so a stale client cannot slip one past it.
A rejected switch restores the previous value and surfaces the host's message
rather than leaving the seat showing something the session is not running.
`agentPreset.select` recomposes one session's agent from a different preset.
It is allowed only while the session is blank — once a turn has run, that
history was produced under the preset's tools and swapping them would strand
logged tool calls, so the attempt answers `agent-preset-locked`.
The agent and the session survive; only the preset subtree is swapped. That
was forced by what the host actually owns: api-proxy discards the `AgentHandle`
it creates, and there is no delete RPC, so neither disposing nor recreating the
session was available. Swapping the subtree is also the better answer — the
session id, its workspace attachment, and its projections all stay put.
`recompose` is unmount-then-mount because two compositions cannot coexist: both
would register the same tool names into one layer. So it resolves the new
preset BEFORE tearing anything down (an unknown id is a no-op) and restores
the previous composition when the new one fails to mount, rather than leaving
the agent with no tools at all. Both paths are pinned by test.
Also restores the English half of the `agentPreset.list` README paragraph,
which was lost before the previous commit — and `verify-translation-pairing
--write` recorded the pair as consistent anyway, because it records whatever
state it finds rather than checking the two sides say the same thing.
One settings row picking which preset new sessions are composed from.
It is deliberately a new-session preference, not a live switch: a session's
preset is fixed at creation and the host refuses to adopt an existing session
under a different one, so the row says "applies to sessions you start from now
on" rather than implying it can retune a running agent.
Options and the current value come from one `agentPreset.list` call — the
roster already reports which id an unspecified session gets, so the row needs
no settings-schema introspection, unlike the permission row it is modelled on.
The write targets only the namespace's `default` field.
The menu marks `user` rows: a locally authored preset is exactly as privileged
as the plugins it names, and presenting it identically to a shipped one would
hide that.
An empty roster reads as `unavailable` and renders nothing, because composing
no presets is a valid deployment rather than a failure — distinct from a
roster call that failed, which surfaces its message.
`agentPreset.list` gives a browser the deployment's roster so it can offer a
choice when starting a session. Each row carries the id, its `trust`, and
whether it is the current default.
`trust` is on the wire deliberately: a `user` preset is exactly as privileged
as the plugins it names, so a surface that offers one alongside a shipped
preset can say which is which rather than presenting both as vetted.
The domain is read-only. A preset is a composition on disk, so authoring one
is a filesystem act rather than an RPC; and a deployment composing no presets
answers with an empty roster rather than an error, because sharing the host
composition is a valid deployment.
The RPC map made every registration site a type error, so the route, the
response-schema table, the service delegate, and the browser fixture are all
wired rather than only the ones I remembered.
knip: ui-settings no longer imports react-dom (the portal moved into
OnboardingSurface), so the react-dom peer/dev dependencies and
@types/react-dom go away with the usage.
Review follow-ups: the Agent Note's Testing section now records the
held-join reload regression scenario this PR adds (both languages,
pairing re-recorded); the e2e hold gates EVERY settings.describe issued
before release instead of only the first, so a future boot-time join
consumer cannot silently collapse the widened window; the sampler's
persistence across later navigations is documented and the vacuity
wording softened to what the hold actually buys (timing independence).
The settings shell painted the onboarding overlay (opaque stage, mask,
#root inert) the moment a step was registered and not locally completed,
while every step still had to load its private join before deciding
whether to show — rendering null could not suppress the shell-owned
chrome. Every reload on the hero therefore flashed a full-screen opaque
layer (white in the light palette) for one settings/credential RPC
round-trip after the session list turned ready.
The chrome now belongs to the step: a new zero-cordis OnboardingSurface
primitive (ui-primitives) renders the body-portaled overlay/mask/stage
verbatim from the former SettingsRoot stylesheet and holds #root inert
for exactly its own lifetime. WelcomeNotice and DeepSeekOnboardingDialog
wrap only their visible branch in it, so their existing null branches
paint and block nothing by construction. SettingsRoot keeps the
coordinator unchanged but renders the elected step bare, and the
settings.onboarding contract now names the surface wrap as the
registrant's obligation.
The onboarding e2e gains a held-join reload scenario pinning that a
configured world never mounts the takeover chrome or inerts the app.
The branch control on user and consumed-steering bubbles could enable
only when a completed turn ended with no node after the message at all,
so readers saw a permanently unavailable control whose tooltip promised
a state it could not reach. An enabled one would mislead anyway: a fork
at a message seq cuts at the containing turn/end and takes the answer
along, the opposite of the branch-to-re-ask reading a control on one's
own bubble suggests.
MessageItem loses its fork props, PendingSteeringBubble loses the
showBranch special case, and messageBranchSeqs narrows to
assistantBranchSeqs: only a completed turn's transcript tail that is
the turn's own content-text assistant may fork. A steered turn keeps
its fork point under the settled answer, because fork is a log-prefix
cut and the steer is model-visible history the child inherits.
Web aria goldens drop the user-bubble disabled-branch row and its
hidden explanation text; the nested-subagent golden also loses the one
enabled user-tail fork handle, a loss the decision note accepts.
Correct the Agent Note's consequence: a running turn withholds the footer
below its own trigger bubble, while every earlier completed turn keeps its
seat — which the package test already asserts.
Give the running-phase barrier an explicit budget: it is armed before the
park and awaited after the stop click, so the 30s replay default left no
headroom for the marker poll, the UI polls, and two aria captures.
Number the running-turn test's boundary seqs like the log does, with each
turn/end strictly after its own nodes.
`assistantActionsSeqs` picked the last content-text assistant of each turn
from the finalized transcript alone. That quantity is stable only once the
turn closes: while a turn is still producing steps, the narration written
before a tool call is the last content assistant so far, so copy, branch,
and the clock appeared under an intermediate sentence for as long as the
tool ran and then moved down to the next step's text.
Pass `ConversationSnapshot.turnEnds` into the derivation and grant the row
only inside a turn that has a durable `turn/end`. This is the same
completion fact the branch control and the `Ran for` label already read, so
the three parts of one row now agree; mid-turn narration owns nothing, and
the seat appears once under the settled answer.
`hasContentText` moves to chat-flow.ts so the ownership gate and
AssistantMarkdown's mount gate cannot drift apart.
apps/web/tests/turn-tail-actions.e2e.ts pins both states through the
assembled application: a hang sidecar on the second model call parks a turn
whose first step narrated before calling bash, and the two goldens hold the
parked flow and the flow after stopping.
Every capacity assertion typed into the field first, so the path that reads a
stored count — the one an already-configured route takes every time it is
opened — was never exercised, and the coverage gate said so. A row now opens
on stored counts and is expected to spell them `1M` and `256K`, the same
vocabulary the field accepts.
`capacityText` takes the row it is rendering rather than looking it up again
by index, which retires the impossible empty-row branch that lookup needed.
The list editor's pre-flight model check is unreachable from the card that
disables submit on the same failure, and says so where it stands.
The two ways to gain a provider had picked up the shared button base's pill
shape and shrunk to their labels, so they read as two stray buttons of
different lengths under the list instead of its last slot. They split the row
evenly again, on the row cards' own corner and the dashed outline this page
already uses for "nothing here yet"; the rule that overrides the base now
says so in one place rather than layering a second `.addButton` block.
An empty capacity shows the adapter's route-level fallback as its placeholder,
so a blank field reads as "sized by the route" rather than as a model with no
capacity. It is a hint, not a mirror: the field counts K as 1000 while the
fallback is 262144, and a deployment may override it.
The picker's description says what the list is without promising an edit the
rows themselves already offer.
The `@media (prefers-reduced-motion: reduce)` block never closed, so every
rule after it — the whole fetch dialog, its candidate list, and the rows
inside — applied only to viewers whose system asks for reduced motion. The
sheet still parsed and the classes still attached, so the list painted with
the browser's own bullets, indentation, and inline label, and the reported
model id ran straight into the capacity beside it.
Report the id alone: it is the string adoption writes, and the capacities the
endpoint disclosed are adopted with it and editable in the row that appears.
The candidate row's remaining font name is the one the theme declares.
The styles gate now checks that the sheet's braces balance, and reads every
theme sheet rather than the platform tokens alone, so a name declared in a
sibling is not called undeclared and a `--dsh-` typo cannot pass as a token.
The Models page could name a provider's key and little else. Adding an
OpenAI-compatible gateway meant opening $DSH_HOME/settings.yaml and
knowing the profile shape; correcting a stale context window meant the
same. This layer puts both on the page: a card that declares a route
pi-ai does not ship — id, endpoint, protocol, key, models — and a model
list on the pi-ai editor that can ask the provider what it serves and
adopt the answer.
It follows the DeepSeek catalog editor that landed in #1050 rather than
inventing a second look for the same job. Both editors now share the
section shell and heading, the danger-tinted delete, the add-model
button, the empty state, the per-row validator that names a bad row by
its position, and one K/M capacity vocabulary — 256K and 1M are read and
spelled back, while settings.yaml still stores plain token counts. The
row type is structurally open like that editor's, so a profile field
this card does not edit survives an edit here.
Three of that editor's decisions replaced weaker ones this branch had
made. Inheritance now reads the composition base rather than the
effective value, which would echo an override back the moment a reset
dropped it. Validation names the offending row instead of stating a
blanket problem. And emptying the list is no longer conflated with
handing the catalog back to the adapter — those are separate acts, with
separate affordances.
The create write carries the revision the card opened at, so a route
another tab declared meanwhile is a conflict rather than a silent
overwrite of its profile.
Restore the replaced pipeline's synthetic fence newline so a real trailing
blank line inside a fence survives CodeBlock's display trim, and pin it
plus header-only tables in the DOM-parity corpus. Route
extractMarkdownPlainText through parseGfm so the grammar has one source.
Document the fixture provenance check (all 46 fixtures reproduce
byte-identically from react-markdown at 9e8101b800), the deliberate
O(prefix) divergence memcmp, the per-arm scope of grammar consistency,
the React 18 MathML namespace limitation, and the prefix-equivalence
corpus constraint.
Port the two master-side markdown fixes onto the mdast renderer:
cjkFriendlyStrong joins both parse.ts grammars as a micromark extension
(ex remark plugin), and inline-code HTTP(S) URLs gain the safe anchor in
render.tsx without normalizeUri. Re-record the two new web-e2e goldens
whose footer clock spacing diverged from the nine established goldens,
and drop the timing-flaky LLM duration segment one of them captured.
