Identify a local review comparison by its compared ref instead of the
displayed base name. Branch records stored before this change appear in
the automatic comparison again, and comparisons that share a name, such
as a worktree start and a local branch, stay apart.
List base branch refs within maxDiffBytes, so a repository with more refs
than maxOutputBytes allows keeps a working menu, and document the
remaining limit.
Move the merge button fill toward the label color so its label reaches
4.5:1 contrast in both themes, name the automatic menu group, and drop an
unused DOM attribute.
Refs #7781
* fix(ui-conversation): rebuild the conversation window after a failed incremental flush
* fix(ui-conversation): keep the original failure when the rebuild also fails
* test(ui-conversation): type the recovery probe view for type-aware lint
* fix(ui-conversation): finish the rebuild inside flush and report it with ConversationUpdateError
* test(ci): retain process identity while awaiting runner recovery
The Branch scope of the local changes view compared only with the
automatic fork point. Its header now names the checked-out branch and a
searchable base branch menu that lists the automatic fork point, then
remote-tracking and local branches; a chosen branch compares with its
merge base with HEAD. The choice lives in the view tab, a missing or
unrelated branch offers the default instead of a retry, and local Branch
review quotes name the base they were compared with.
Refs #7781
* fix(session-projection): keep released units in checkpoints
Since the Cordis rc.11 teardown order unloads the Web layer before the
base layer, Session Controller and harness-dev-context release their
projection units before their live Sessions detach. The projection
cache is still alive at that point and wrote a checkpoint holding only
the units that were still registered, replacing the complete record.
A blank Session then lost sessionListMetadata and listed as an untitled
conversation after every graceful restart.
A released unit now stays in the registry as a predecessor: it leaves
event drives, snapshots, restores, and change notifications, while
checkpoint() keeps each Session's last cell of it without advancing it.
Registering the same key again replaces the predecessor and refolds
the log. The projection cache keeps replacing whole records.
* test(session-projection): assert frozen released units across unload tests
Update the remaining unload tests that still expected a released unit
to vanish from checkpoint(): user-approval, agent-loop inbox and scope
lifecycle, time-context, and the agent registry now assert that the
same Session keeps its pre-unload row, that a later event does not
advance it, and that a Session created after the unload has no row.
The token-meter and goal unload tests also append an event after the
unload.
Document that a retained handle read still advances a released cell
that checkpoint() serves, and that restore() folds only live units, so
its refreshed rows omit released units.
* docs(session-projection): sync the ProjectionHandle type-equiv block
* docs(session-projection): record frozen released units and their upgrade
Add the upgrade guide for checkpoint() keeping a released unit's last
row, update the registration-handles Agent Note with the frozen-cell
decision and the retained predecessor, list both consequences under the
registry README's known limitations, note in the projection cache README
that a stored row can be frozen at its unit's release, and say that
restoreFloor() ignores released predecessors.
Update the agent-loop resource-lifecycle unload test, which read
checkpoints through an alias, to the frozen-row semantics.
* docs(upgrade-guide): mark the released-checkpoint snippet ignore-check
* docs(session-projection): state the retained-handle path and narrow a rejected alternative
* docs(upgrade-guide): re-record the released-checkpoint guide pairing
* feat(web): boot past failed optional client plugins
The Web client boot audit now follows Host startup strictness. An
inactive entry on the global required list (module system, Session
transport, renderer, application frame, Session view, and sidebar)
still rejects startup and the report marks the required entries.
Other inactive entries produce one console warning and the
application mounts without them. The boot page no longer draws a
failed entry before the audit decides.
* fix(web): audit boot entries whose creation rejects and name activation errors
ClientEntries.start() now logs a row whose Loader creation rejects, such
as a module that exports no plugin, and leaves it inactive for the boot
audit, as the Host Loader group does. The audit awaits only failed
fibers to report their recorded activation error. Notes, READMEs, and
the Web client subsystem page describe the browser boot audit; tests
cover invalid exports, carrier presentation of optional failures, and
required names absent from the boot graph.
* test(web): fail test carriers and e2e on inactive client entries
The product boot only warns about inactive entries outside its required
list, but tests must still surface every plugin that fails. TestClient
and createClientTest now require every roster entry to activate unless
a case passes allowInactiveEntries. The watchConsole tripwire, the
real-host smoke, and the assembled jsdom boot treat a web boot warning
as a failure with no opt-out.
Docs: the Web README names Session control, notes that an entry without
a fiber is reported as an import failure like the Host audit's "failed
to import", and the startup strictness note lists the browser tests.
Review round on the naming change.
`resolveName` now distinguishes an address from a Session id: a
slash-separated address never is one, so an unmatched path fails with
AGENT_NAME_NOT_FOUND instead of reaching member lookup as an unknown id,
while a single token keeps the Session-id fallback. Path resolution also
treats external and unknown children as leaves, so a name below an external
run reports the same clear failure instead of surfacing a session-query error
for an identity that owns no local Session.
Native V4 admission rejects a repeated sibling name in own catalog facts,
symmetrically with its duplicate-child check, so a log that would make name
lookup ambiguous fails before any member is resolved.
`SiblingNameReservations.release()` checks that it still owns the map entry,
so a stale or repeated release cannot drop a newer reservation for the same
creator.
`list_agents` explains corrupt, unsupported, and unavailable: they report that
the row's own descendants could not be read, and the row remains a member the
caller can message. The earlier wording restricted messaging to running or
inactive rows, which excluded addressable members.
`isAgentName` stays module-local; the package entry exports only
ROOT_AGENT_NAME, matching its current callers.
Package READMEs, the subsystem page, the Agent Note, and the generated API
catalogue follow; recorded-session, CLI, and Web expectations are refreshed.
The `ctx.mcpServers` key reads as a collection of servers, but the
service manages that collection, and ACP already uses `mcpServers` for
a plain server list. Publish the service as `ctx.mcpManager`, matching
the package name, the `mcp-manager/` event prefix, and
`ctx.pluginManager`. Remote endpoints become `mcpManager/*`; the class
keeps its name, McpServerManager.
Models coordinate subagents through send_message, interrupt_agent, list_agents,
and Team task owners, but every one of those addressed a member by a
36-character Session id. Each new child now has a name, unique among its
creator's children, recorded as a field of the subagent/catalog entry the
creator already appends (payload version 3).
The Session id stays the only execution identity for routing, authorization,
persistence, task ownership, and navigation. A name is a model-facing
reference: 1-32 lower-case letters, digits, or hyphens starting with a letter,
never `lead`, so a UUID never parses as one. Omission selects the first free
`agent-N`; an explicit duplicate or invalid name fails before startup. The
creator reserves the name synchronously at admission and releases it when the
catalog entry commits or startup fails, so no second write, rollback protocol,
or optional naming service exists.
The root is `lead`, and a descendant's full name is its path from the root
without a `lead/` prefix, such as `reviewer/checker`. Sibling-unique segments
make full names tree-unique without a root registry, and cold resume rebuilds
a member's path from the catalogs along its creation chain. A member below an
unnamed historical creator keeps its Session id as its name. resolveName()
reads only the catalogs along the path and grants no authority.
Tools, runtime-authored text, and the Web client consume names: delegation
takes an optional name and returns { kind, subagentId, name }; send_message and
interrupt_agent take `agent`; list_agents rows carry name, status, and
description; Team task results report the owner's name while storage keeps
ownerId; identity runtime context, message framing, and settlement notices use
full names. Recorded sessions, CLI and Web expectations, the Python runtime
snapshots, bilingual docs, an upgrade guide, and an Agent Note follow.
* perf(session): send one projection frame per committed event
* fix(session): deliver keys changed before a drive failure; address review
* perf(session): carry projection frame changes as key and val entries
Xu Hanxiang