Three things the page got wrong.
The web-search provider's defaults lived only at their use site, so the
served section carried no value for them and the card fell back to a zero it
invented — a number the schema itself rejects. Declaring them on the schema
makes the settings service the one authority: `maxUses` now reads 5 because
that is what the Host resolves, not because the page guessed. `baseURL` keeps
its code-side default, which exists so `$DEEPSEEK_SEARCH_BASE_URL` can win.
A field the Host serves no value for now renders empty rather than as zero.
The cards were rows in a settings page of cards: name and description ran
together on one line because the shared disclosure row lays them side by side.
Each card now draws its own header, stacking the two, and the section follows
the idiom the Agent Preset page established.
A restriction was compiled against the global tool layer alone: only
global-layer tools were tested against `admits()`, and every chain-layer
tool was overlaid unfiltered afterward. That read the exempt set as "the
global layer" when what it means is "what this scope registers itself" —
two descriptions of the same set only while every model-facing tool sat in
the host composition.
Moving those rows onto the agent plane separated them. A preset's tools are
an ANCESTOR contribution to a joined agent, so a subagent's `toolFilter`
stopped constraining anything it was given; and with the global layer empty
`restrict()` rejected every name it received as unknown, failing the child
outright. With the same tools in the global layer the filter still admits
and applies normally, which is what makes this a regression of the move
rather than a standing limitation.
`view()` now filters everything a scope inherits — the global layer and
every ancestor layer on its chain — and exempts only the layer the scope
owns. That exemption is load-bearing rather than incidental: the delegation
runtime registers a child's `report` and structured-output tools into the
child's own layer, and a filter naming the capabilities the child may use
must not strip the machinery it answers through. Tool order, and with it
prefix-cache reuse, is unchanged: inherited names keep their global-then-
ancestor position and own-layer names still come last.
The diagnostic said "unknown global tool" while listing what is really the
inherited surface; it now names the surface it checks and says why an
own-layer name is not restrictable.
Fixes#2185
The locale plugin now derives its durable preference from a settings scope
bound to the connection service, so the trajectory benches provide a
connection handle before loading it.
A per-instance connection supervisor restarts the original server config
with exponential backoff when the transport closes, re-runs tool discovery
on success, and atomically replaces the previous generation. Default policy
retries for ~2.5 minutes (10 attempts, 500ms→30s doubling) before giving up
and unregistering the server's tools.
New config block reconnect { enabled, initialDelayMs, maxDelayMs, maxAttempts }
on both transports; misconfiguration fails plugin load. A connection that
survives past the stability window (maxDelayMs) resets the attempt budget,
so occasional crashes recover indefinitely while a crash loop still exhausts
the cap.
Integrates with the upstream failOnStartupError: the initial sync uses
registrationFailure:'throw' when that flag is set so a squatted namespace
still rejects activation.
Fixes#1746
Also folds the duplication the three controls had grown: the draft-and-commit
input both editable fields render is now one component, and the two sibling
executors' identical import surface is marked as the deliberate mirror their
READMEs already describe.
The persistence subsystem page gains the SessionRawArtifact section (with a
type-equiv manifest entry and zh counterpart), and the README and Agent Note
describe the drain-based backpressure instead of claiming the archive never
accumulates.
Toolbar strings route through the locale dictionary's standard t seat (the
export button no longer mixes languages in the English golden), exportLog
passes the response blob straight to the browser save instead of copying it
three times, the client id sanitizer rejects dot segments like the host one,
and the fixture stub comment no longer misattributes the 404.
Chunk boundaries never split a surrogate pair (a lone high surrogate
re-encodes as U+FFFD and silently corrupts the exported artifact), production
yields whenever the response queue fills so a slow consumer bounds the
accumulation, includeDescendants rejects values other than true/false instead
of silently under-exporting, the dead missing-services arm is deleted by
narrowing the streaming deps, and the readRaw failure answers 500 without
leaking host paths into the browser error bar.
The inherited readRaw default now rejects like the async backend overrides
instead of throwing synchronously, and its arms plus the JSONL override's
retry loop, zero-frame, and corrupt-header branches get dedicated tests.
