| Age | Commit message (Collapse) | Author |
|
|
|
# Conflicts:
# backend-handoff.md
# src/features/workspaces/ui/WorkspaceCard.test.ts
|
|
# Conflicts:
# backend-handoff.md
|
|
|
|
|
|
Consumes the backend's concurrency-fixes (commit 2d27666) — additive to
[email protected], NO version bump. ConcurrencyStatusEntry gains
cooldownMs / autoReduced / autoReducedFrom? / notice?; new
ConcurrencyCooldownResponse / SetConcurrencyCooldownRequest + GET/PUT
/concurrency/cooldown/:providerId.
FE:
- Pure core (logic/view-model.ts): DEFAULT_COOLDOWN_MS; parseCooldownInput
(non-negative int — 0 is valid, unlike the limit); normalizeCooldown;
cooldownLabel ("350ms"/"1.2s"/"0ms (off)"); viewConcurrencyStatus extended
(cooldown + autoReduce fields; auto-reduce → warning badge, not busy);
viewAutoReduce/autoReduceNotices (banner view — prefers backend notice,
synthesizes a fallback); summarizeStatus "N auto-reduced" fragment;
normalizeConcurrencyStatus coerces new fields (immutable readonly build;
autoReducedFrom/notice only when autoReduced===true); normalizeConcurrencyCooldown.
- Types (logic/types.ts): re-export the 2 new contract types +
ConcurrencyCooldownResult + Get/SaveConcurrencyCooldown ports.
- UI: ConcurrencyCooldownRow.svelte (inline-edit cooldown + Save → PUT, seeded
via the ChatLimitField pattern); AutoReduceBanner.svelte (dismissible banner —
backend notice + "Was N, now M." + "Restore to N"); ConcurrencyView renders
the cooldown per status card + the banner section. The banner persists while
autoReduced===true, is dismissible, and clears automatically once a poll shows
autoReduced===false after a restore PUT.
- Store (store.svelte.ts): getConcurrencyCooldown + setConcurrencyCooldown
(surface 400/404/503 as ok:false; normalize at the seam) + interface decls.
- App.svelte: saveConcurrencyCooldown adapter → ConcurrencyView.
- Tests: +47 (view-model cooldown/auto-reduce/normalizers; component banner
render, cooldown PUT, restore clears banner, dismiss; store cooldown GET/PUT).
Re-synced the file: dep (bun install) + re-mirrored .dispatch/transport-contract.
reference.md. typecheck 0/0, 1048 tests green (run twice), biome clean, build OK.
Worktree env: an untracked dispatch-backend → backend symlink was created in the
worktree parent so the canonical file:../dispatch-backend/... paths resolve
(NOT committed — per the §2d/§2j worktree convention).
|
|
active chats
|
|
Backend (feature/workspace-star) shipped Workspace.starred: boolean (additive
to [email protected], no version bump) + PUT/DELETE /workspaces/:id/star endpoints
(no body; create-on-miss; return the updated Workspace). A starred workspace's
agents jump ahead of non-starred ones in the concurrency limiter queue
(oldest-agent-first within each group); takes effect immediately for
already-queued agents.
FE consumed:
- adapter/http.ts: star(id)/unstar(id) -> WorkspaceResult<Workspace>
(PUT/DELETE /workspaces/:id/star, no body).
- logic/view-model.ts: pure sortWorkspaces (starred-first, then lastActivityAt
desc, stable) + pure applyStarred (the optimistic apply/revert transform).
- store.svelte.ts: setStarred(id, starred) — optimistic flip with error revert;
the list is now a $derived sorted view (starred bubble to top reactively);
no full refresh on success (avoids flicker).
- ui/WorkspaceCard.svelte: star toggle button (filled gold when starred,
outline when not; spinner in flight; aria-pressed/aria-label; tooltip notes
concurrency priority).
- Re-mirrored .dispatch/wire.reference.md (starred + delta note).
- GLOSSARY.md: 'starred' term.
Tests (+27): http star/unstar (5), view-model sort+applyStarred (12), store
optimistic+revert+re-sort (6), WorkspaceCard star button (4).
Verification: typecheck 0/0, 1045 tests green, biome clean, build OK.
backend-handoff.md updated (workspace-star slice, no open backend asks).
|
|
# Conflicts:
# .dispatch/transport-contract.reference.md
# backend-handoff.md
# src/app/App.svelte
# src/features/chat/ui/Composer.svelte
|
|
Images are now stored on disk under tmp (not SQLite) and served via
GET /images/:conversationId/:imageId. Persisted ImageChunk.url is a compact
relative HTTP path (/images/<conv>/<uuid>.png) instead of a base64 data URL.
No wire/transport-contract type change (behavior only) — re-mirrored the
delta notes.
- New pure resolveImageUrl(url, apiBase) helper (core/chunks/image-url.ts,
+8 tests): data/absolute URLs pass through; relative paths are prepended
with the API base (no double slash; empty base -> root-relative). Exported
from core/chunks + re-exported from features/chat.
- ChatView: new apiBaseUrl prop; <img src> uses resolveImageUrl. The
optimistic echo's data URL passes through; persisted relative paths
resolve against the base. +3 tests.
- AppStore exposes httpBase (getter); App.svelte passes apiBaseUrl into
ChatView and the heartbeat RunModal (also renders image chunks).
Verification: svelte-check 0/0; vitest 959/959 (run twice, +11); biome
clean; vite build OK. See backend-handoff.md §2j-update-2.
Not merged or pushed.
|
|
Backend now persists concurrency limits across reboots (no API contract change
— same endpoints/shapes, so no re-pin/re-mirror needed). Taking the suggested
optional UX hint: after a successful Save the limit row shows a brief green
'Saved.' that clears on the next edit (mirrors the ChatLimitField pattern).
backend-handoff.md §2j: notes the persistence change + the two earlier
backend-only changes (200ms release cooldown; 'queued' status — already shipped
as CR-13). typecheck 0/0, 926 tests green, biome clean, build OK.
|
|
The 'Concurrency limits' and 'Live status' section headings are clear on their
own; the explanatory paragraphs under each were overbearing. Removed both.
|
|
The 'queued' (waiting-for-a-concurrency-slot) indicator on the tab + composer
corner used loading-spinner; switch to DaisyUI's loading-ring. The 'active'
(generating) indicator stays loading-dots. typecheck 0/0, 926 tests green,
biome clean, build OK.
|
|
The 2s status poll (and post-mutation limit reloads) toggled a visible
loading state, but since the refresh is near-instant the spinner flashed
for <1 frame — a visible flicker/jerk every poll. The polling is now
SILENT (mirrors the heartbeat runs list):
- refreshLimits/refreshStatus: re-entrancy guard (limitsInFlight/statusInFlight,
no UI), no loading-state toggle. Error cleared only on success so it stays
visible + stable during an in-flight retry (no vanish-mid-poll).
- Removed limitsLoading/statusLoading: the Refresh buttons are plain-text
(no spinner, not disabled); the empty-state guards use hasLoaded* only
(no && !loading), so the list area never reflows mid-refresh.
+1 regression-guard test (Refresh buttons never surface a spinner). typecheck
0/0, 926 tests green (x2), biome clean, build OK.
|
|
Backend vision update (additive to [email protected] / [email protected],
no version bump).
Contracts mirrored (.dispatch/transport-contract.reference.md):
- VisionSettingsResponse + SetVisionSettingsRequest (GET/PUT /settings/vision).
- Delta note: read_image -> consult_vision; numbered placeholders; compaction.
Tool rendering (ChatView):
- read_image test -> consult_vision (rendering is generic by toolName).
- +2 tests: numbered-placeholder text chunk + [Compacted image] text chunk
(both regular text chunks, render as-is — no special handling).
New vision feature library (src/features/vision/):
- logic/view-model.ts (32 tests): VisionSettings/VisionSettingsPatch types
(consumer-defines-port), LoadVisionSettings/SaveVisionSettings ports +
results, normalizeVisionSettings (network-seam coercion), parseImageLimit/
imageLimitChanged, compactionModelOptions (vision-capable models via chat's
public isVisionModel + Auto sentinel), round-trip helpers, imageLimitLabel.
- ui/VisionSettingsView.svelte (9 tests): imageLimit input + Save,
compactionModel dropdown (Auto + vision-capable models), load-on-mount,
save-on-change, error/saved feedback.
- index.ts.
Cross-unit seam: isVisionModel added to features/chat public index.ts
(additive); imported through the public surface, not internals.
Store wiring (src/app/store.svelte.ts):
- visionSettings state + refreshVisionSettings (GET /settings/vision,
normalized at the seam) + setVisionSettings (PUT, partial, returns merged) +
VisionSettingsResult; seeded on boot; exposed on AppStore. +4 store tests.
Mounted in App.svelte: new "Vision" sidebar view kind + VisionSettingsView
in viewContent (not conversation-scoped); load/save adapters; visionManifest.
Verification: svelte-check 0/0; vitest 948/948 (run twice, +47 since the
prior vision commit); biome clean; vite build OK. See backend-handoff.md §2j.
Not merged or pushed.
|
|
Backend shipped "queued" ConversationStatus (additive to [email protected]): when a
request blocks on a concurrency slot, conversation.statusChanged broadcasts
"queued" (broadcast-only, never persisted); "active" on slot grant.
FE consumes it:
- WS parser (adapters/ws/logic.ts): accepts "queued" in the status set.
- Store handler: "queued" updates the status map + opens a tab for a new
cross-device queued conversation (like "active").
- TabList: status === "queued" -> loading-ring (spinner, aria-label "Queued");
"active" -> loading-dots (unchanged).
- Composer: status type widened to ComposerStatus (idle|running|queued|error),
exported from features/chat. "queued" -> a loading-ring status icon +
placeholder "Queued for a slot…"; behaves like "running" for the send
button (steer/stop — the turn is in flight, just waiting for a slot).
- App.svelte: composerStatus derived (error > queued > running > idle) —
conversationStatus === "queued" wins over generating so the corner shows a
ring during the wait (turn-start fires before the slot is granted, so
generating is already true while status === "queued").
- Re-mirrored .dispatch/wire.reference.md (ConversationStatus widened + header).
Tests: WS parser accepts queued; store handler sets status + opens a cross-device
tab + transitions queued->active->idle; TabList renders a ring for queued + dots
for active. typecheck 0/0, 925 tests green (x2), biome clean, build OK.
backend-handoff.md CR-13 marked RESOLVED.
|
|
|
|
Replace the Add-form provider text input with a <select> dropdown populated
from the available models' provider prefixes (<provider>/<model>), unioned
with providers already carrying a configured limit (so a limit set out-of-band
but whose model list is empty still appears). The selection auto-defaults to
the first option and falls back when an option disappears.
- logic/view-model.ts: pure providerFromModel + providerOptions(models, limits)
(distinct provider ids, first-seen order) + 7 tests.
- ConcurrencyView.svelte: models prop + <select> bound to newProviderId;
disabled + 'No providers available' when there are no models.
- App.svelte: pass models={store.models}; component test updated to the
dropdown (selectOptions) + a no-models case (6 component tests).
Verified: typecheck 0/0, 922 tests green, biome clean, build OK.
|
|
|
|
against sidebar
|
|
|
|
swapping text
|
|
|
|
Vision & vision-handoff frontend (consumes the backend's additive
[email protected] / [email protected] image types — no version bump).
Contracts mirrored:
- .dispatch/wire.reference.md: ImageChunk added to the Chunk union +
ImageChunk/ImageInput interfaces.
- .dispatch/transport-contract.reference.md: ChatRequest.images,
ModelMetadata.vision, + ImageChunk/ImageInput re-exports.
Core (core/chunks):
- conformance: assertChunkExhaustive handles the new 'image' variant
(the guard caught it — its purpose).
- appendUserMessage(state, text, images?) echoes a [text, image, ...]
user run; the user-message event dedup scans the trailing user run
(not just the last chunk) so an image-bearing echo doesn't duplicate
the text; applyHistory's during-gen dedup matches a multi-chunk echo
by content equality (chunkContentEquals + trailingRun helpers).
UI:
- ChatView renders user 'image' chunks as lazy <img> bubbles; a
non-vision model's persisted [image, analysis-text] both render.
read_image tool renders generically (no special-casing).
- Composer: clipboard paste / file picker / drag-drop of images ->
base64 data URLs, thumbnail previews with remove, forwarded on
chat.send (omitted when none). Image-only sends allowed; steering
(chat.queue) never forwards images.
- ModelSelector: vision badge (isVisionModel) marks vision-capable
models; indicator shows native-vision vs vision-handoff hint.
Store wiring: ChatStore.send + AppStore.send + App.svelte handleSend
thread images through; chat.send still omits cwd (only images added).
Verification: svelte-check 0/0; vitest 901/901 (run twice, +34 new);
biome clean; vite build OK. See backend-handoff.md §2j.
Not merged or pushed.
|
|
- backend-handoff.md: full §2j slice (API surface, contract note, FE summary,
verification, the single-provider GET note, worktree symlink note) + bump the
pinned transport-contract version 0.22.0 → 0.23.0 in the header/packages/
mirror/endpoints sections.
- ConcurrencyView.svelte: remove an unused ConcurrencyLimitView type import
(the type flows via viewConcurrencyLimits' return type; svelte-check clean).
|
|
|
|
|
|
# Conflicts:
# src/app/App.svelte
# src/app/store.svelte.ts
# src/app/store.test.ts
# src/features/workspaces/ui/WorkspaceCard.test.ts
|
|
endpoint
|
|
|
|
|
|
|
|
|
|
|
|
|
|
sidebar wiring
|
|
# Conflicts:
# src/features/workspaces/ui/WorkspaceCard.svelte
# src/features/workspaces/ui/WorkspaceCard.test.ts
# src/features/workspaces/ui/WorkspacesHome.svelte
|
|
When the user clicks Test, the button now transitions spinner → ✓ (green) on
success / ✗ (red) on failure (persisting until the next test or computer
change), instead of reverting to plain 'Test' text with no success indicator.
Also refresh the connection-status badge immediately after a test completes, so
a stale 'connecting' spinner caught by the 4s poll mid-test is cleared and the
badge reflects the real post-test state. Clear the test result when the selected
computer changes so a stale ✓ from alias A doesn't persist for alias B.
|
|
Mirrors the cwd/workspace UI for the SSH-computer feature:
- New feature library src/features/computer/:
- logic/view-model.ts (pure): viewComputer/viewComputerStatus/viewTestResult/
summarizeComputers/formatHost/knownHostLabel + state->badge for the 4
ComputerStatusResponse states + SaveComputer/LoadComputerStatus/
TestComputer/LoadComputers ports. 20 view-model tests.
- ui/ComputerField.svelte: per-conversation selector (dropdown +
connection-status badge + Test-connection, polling the selected alias).
- ui/ComputerSelect.svelte: reusable Local/computers dropdown, shared with
the workspace default-computer control.
- AppStore: computerId state + refreshComputer (at every focus site, parallel
to refreshCwd) + setComputer (PUT /conversations/:id/computer, null=clear) +
global computers catalog (GET /computers on boot, like models) +
computerStatus(alias) + testComputer(alias). chat.send UNCHANGED (resolved
server-side like cwd).
- App.svelte: ComputerField in the Model sidebar view next to CwdField; adapted
ports wrap the store.
- workspaces: setDefaultComputer on WorkspaceHttp+WorkspaceStore
(PUT /workspaces/:id/default-computer); default-computer selector in
WorkspaceCard (reuses ComputerSelect); router passes store.computers through.
- Re-mirrored .dispatch/transport-contract.reference.md (Computers section +
ChatRequest.computerId); updated .dispatch/wire.reference.md (Computer/
ComputerEntry/defaultComputerId + the provider-retry divergence note from
handoff #1); GLOSSARY + backend-handoff.md (handoff #2, §2e).
Transparency invariant: the computer is USER-facing only (a tool-execution
target, never part of the model prompt); the agent never sees it.
Verify: 795/795 tests green; biome clean; vite build succeeds; 0 typecheck
errors from the computer feature. (11 pre-existing svelte-check errors remain
from the open §2d provider-retry divergence — backend feature/ssh-support
still lacks TurnProviderRetryEvent; not from this feature.)
|
|
The 'Open' button on a workspace card did client-side navigation
(onNavigate -> pushState), opening the workspace in the SAME tab.
Replace the JS click handler with a native anchor using target="_blank"
(+ rel=noopener noreferrer) so clicking Open opens the workspace in a
NEW browser tab. The new tab does a full load and routes to /<id> via
the existing parsePath -> App wiring.
The card no longer needs the onNavigate prop (its only navigation was
the Open button), so drop it from WorkspaceCard + WorkspacesHome and
update the card tests to assert the new-tab link attributes instead of
the onNavigate call.
|
|
The conversation tab's title span was truncated with a static
'Double-click to rename' tooltip. Replace the static title with the
full tab.title so hovering reveals the complete conversation title
(the truncate class still clips the visible label).
|
|
features + app wiring
- workspaces: URL-driven conversation grouping (home listing at /, routing,
store, http adapter, WorkspaceCard) wired into the App.svelte shell
- rename features/workspace -> features/cwd-lsp (the cwd/lsp status feature)
- new features: mcp (status view), settings (chat-limit field), system-prompt
(prompt builder), all rendered via the generic surface host
- chat: store + ChatView updates
- tabs: tabs-store updates
- app wiring: ErrorModal (full-screen error surface), app/App.svelte + store.svelte
This commit makes HEAD typecheck clean for the first time: the prior HEAD
(dfe1711) imported features/settings from app/App.svelte but never committed
the feature, so only the full working tree was green.
|
|
Double-click a tab's title to enter inline edit mode. Enter or click
away (blur) saves; Escape cancels. The rename is optimistic — the
local tab updates immediately and PUT /title fires in the background.
683 tests green.
|
|
1. Real context window: GET /models now returns modelInfo[model].contextWindow.
The Composer uses this instead of the hardcoded MAX_CONTEXT = 1,000,000.
Falls back to 1M when modelInfo is absent or the model has no contextWindow.
2. Percentage-based auto-compact: the compact-threshold endpoint is renamed
to compact-percent. The CompactionView now shows a percent input (0-100,
default 85, 0 = manual) instead of a token count input. Types renamed:
CompactThresholdResponse → CompactPercentResponse,
SetCompactThresholdRequest → SetCompactPercentRequest.
Note: the field name in the backend types is still 'threshold' (not
'percent') — the FE maps between them.
Re-mirrored .dispatch/transport-contract.reference.md.
686 tests green. 0 svelte-check errors + warnings.
|
|
Sidebar panel layout (which views are open and their order) and the
sidebar open/closed toggle are now persisted to localStorage. Default
layout is just the Model view at the top.
- ViewSidebar accepts an onChange callback that reports panel kinds
- App.svelte creates two createLocalStore instances (dispatch.sidebar.views
+ dispatch.sidebar.open) using the store's storage adapter
- AppStore exposes its storage instance so the shell persists via the
same adapter (test-injectable, not globalThis.localStorage)
- Tests pre-populate fake storage with ["extensions"] for the 4 tests
that need the Extensions view visible
686 tests green. 0 svelte-check warnings (2 pre-existing errors from
missing transport-contract exports, unchanged).
|
|
Tool calls and results now use the same DaisyUI collapse pattern as
thinking blocks — collapsed by default, click to expand. Each card
shows the tool name + a wrench icon in the title; expanding reveals
the input/output with overflow-x-auto for long lines and max-h-96
overflow-y-auto for very long output.
Batched tool calls: each entry is its own collapse card (was a DaisyUI
list). Pending results show a spinner in the title. Errors show a red
badge.
686 tests green.
|
|
Add overflow-hidden to tool card containers and overflow-x-auto to
<pre> elements so long tool output (file contents, JSON, etc.) gets
its own scrollbar instead of expanding the chat width and creating
a horizontal scrollbar on the entire transcript.
686 tests green.
|
|
trimTranscript now drops oldest provisional chunks (the in-flight turn)
when committed chunks are exhausted. Previously it bailed with drop=0
when committed was empty, allowing unbounded provisional growth during
long generating turns (300+ chunks → browser crawls).
Root cause of the syncTail approach failing: the kernel emits
step-complete (line 360) BEFORE calling onStepComplete (line 542) —
chunks are persisted only after tool results come back, not when
step-complete fires. So syncTail on step-complete found nothing.
Reverted the applyHistory + syncTail-on-step-complete changes from
d354626. The new approach is simpler: trim provisional directly in
trimTranscript. Dropped chunks are lost temporarily (no Show Earlier)
but come back as committed when the turn seals and syncTail fetches
everything from the server.
686 tests green.
|
|
The backend now persists chunks at step boundaries during generation
(CR-6). The FE calls syncTail on each step-complete event to fetch
the newly committed chunks. applyHistory clears the provisional array
when new committed chunks arrive mid-generation (they're duplicates
of what was folded from live events). The accumulating chunk (current
in-progress step) is kept.
This means trimTranscript can now drop oldest committed chunks
uniformly during a long turn — no unbounded provisional growth. The
browser never holds more than chatLimit chunks, even mid-generation.
3 new tests: clears provisional on new committed during generation,
keeps provisional when no new chunks, keeps accumulating when clearing.
689 tests green.
|
|
The turn number comes from the entry's position in the metrics array
(1-based), which is correct regardless of trimming since stepId matching
aligns segments to the right entry. Now displays 'turn 3 · 12k tok' instead
of just 'turn · 12k tok'.
|