The structural T-403 child: make vim normal mode mean navigation in panes
that were mouse-only. The passive global key path can't run multi-chord
sequences (D-82), so each pane hosts its own SequenceMatcher — factored into
a reusable PaneKeyNav that resolves the live keymap and dispatches nav.*
intents while a pane holds focus under the vim preset.
- nav.* intents (down/up/pageDown/pageUp/top/bottom/expandOrRight/
collapseOrLeft/activate) — preset-neutral; vim.yaml binds j/k/ctrl+d/ctrl+u/
gg/G/l/h/[o,enter] under `vim.normal && !editor.focused`.
- The editor publishes an `editor.focused` scope flag from its focus node, so
the same keys stay buffer motions while the editor is focused and become nav
when a pane is — resolved by file order + the guard (no change to the editor
motion bindings).
- File tree: a flattened visible-index selection cursor in FileTreeController
(j/k move, h collapse-or-out, l expand-or-into, o/enter open), with a focus
ring + scroll-into-view.
- Conversation: j/k line-scroll, ctrl+d/u half-page, gg top, G bottom — G
re-arms follow-tail.
Foundation for T-404/T-405/T-407, which build on the per-pane matcher and the
editor.focused guard. Git panel + ticket board list nav deferred to a
follow-up (the ticket says lists can trail). Tests: keymap resolution under
both scopes, PaneKeyNav dispatch, the controller selection model, and
end-to-end key-driven nav in both panes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A `Workflow` tool-use launches its multi-agent run in the background and
returns immediately; the real fan-out arrives out-of-band on stream-json
`type:"system"` task_* events (task_started / task_progress / task_updated /
task_notification) keyed by the launching tool-use id — which clide was
dropping. (Wire shape captured by two live stream-json probes; recorded on
the ticket.)
- workflow_run.dart: a pure, Flutter-free WorkflowRun/WorkflowAgent model
that folds those events (phases, per-agent start→progress→done deltas,
usage) into a snapshot.
- StreamJsonSession recognises the events, accumulates a
Map<toolUseId, WorkflowRun>, and exposes `workflows` + `workflowsStream`.
- A `Workflow` tool-use with a live run renders a dedicated run card —
phase groups, per-agent rows with spinner/check status, usage, and the
script — falling back to the generic tool card pre-progress or on reload.
The run breaks the activity cluster so it's always first-class (like T-342).
- The sidebar Activity tab adds a WORKFLOWS section: one row per run with its
done/total agent count, tinted by running/done state.
Closes T-416 and the T-410 epic (all children done).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
clide forwards composer input to a headless (stream-json) CLI, where the
TUI's interactive commands don't exist. A known-but-TUI-only command
errored raw ("/x isn't available in this environment", rendered as fake
claude prose); an un-advertised one (e.g. /effort on 2.1.175) was worse —
bracket-pasted to the model as literal text, burning a real turn.
Probed claude 2.1.175 for ground truth: the initialize handshake's
slash_commands advertises skills + the headless builtins only; forwarded
local-command output comes back as an assistant message with model
"<synthetic>"; set_effort is not a control subtype; /usage works headless.
- slash_commands.dart: SlashRoute routing table (owned > advertised >
TUI-only catalog > forward) + kTuiOnlyCommands with clide-native hints
+ tuiOnlyNotice(). One source of truth replacing ad-hoc checks.
- claude_pane._send routes 'unavailable' to a local notice card; nothing
reaches the session.
- transcript_reader: AssistantTextMessage.synthetic ("<synthetic>" model)
so CLI-local output is distinguishable; "<synthetic>" no longer
clobbers the tracked model in SessionStatus (latent /usage bug).
- conversation_view: synthetic output renders as a muted framed "clide"
card (T-306 styling), never coral Claude prose.
- kFallbackSlashCommands trimmed to the genuinely-headless builtin set —
it doubles as the router's advertised fallback, and the old list's
TUI-only entries would have routed to a raw CLI error.
Board (rides this commit): T-414 gains the user's sidebar styling-pass
note; T-416 filed — surface Claude Code Workflow runs in convo/status.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Wire the detection + follower core into the Bash tool card. A Bash card
with a follow intent (`tail -f …`) gains a "live tail" segment below the
result: an embedded read-only TerminalView fed by FileTailFollower on the
file the command follows, resolved against the open workspace.
Lazy lifecycle for free: the collapser builds its children only when
expanded (clide_collapser_card.dart), so _BashLiveTail starts the follower
in didChangeDependencies on expand and stops it in dispose on collapse —
no follower runs until the card is expanded. No resolvable file-backed
source → a muted "no independent source to follow" note, never an empty
terminal. The workspace root comes from kernel.project.current, so no new
plumbing through the conversation widget tree.
Tests: a tail Bash card surfaces the segment (+ the muted note when no
project/source); an ordinary `ls` card gets no segment; the segment only
builds on expand.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A fan-out of N agents (Task/Agent) merged into one shared "Activity / N
steps" cluster — groupConversation folded an Agent spawn like any Bash/
Read call. Now an Agent spawn is a cluster boundary, rendering as its own
first-class collapsing card (reusing the existing sticky-agent path: folded
prompt T-263 + nested run T-264), while adjacent non-agent foldables keep
clustering into the normal Activity card.
Two changes:
- activity_cluster: a shared isAgentTool() predicate; _isFoldable returns
false for agent spawns at every level (incl. L3), so parallel agents
never merge. Only the grouping boundary changes; fold mechanics are
unchanged.
- conversation_view: harden resolveOwner. Its nearest-preceding-agent
fallback is safe with one agent but mis-routes under a parallel fan-out
(an unattributable item lands in whichever agent was emitted last —
a sibling's card). With >1 agent, drop the fallback so the item orphans
(rendered inline) instead of cross-attributed. The T-338 direct route
(parent_tool_use_id) still attributes interleaved items correctly.
Tests: two consecutive agents → two cards (not one cluster); agent breaks
a sibling cluster; agents first-class at L3; regression — consecutive
Bash still one cluster; interleaved parallel-agent runs route to their own
card; an unattributable item orphans instead of being swept into the last
agent.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Raise the declared minimums in pubspec.yaml to what our deps already
require: Flutter >=3.35.0 / Dart >=3.9.0 (was 3.19.0 / 3.5.0). alchemist
0.12 needs Flutter 3.32; Dart 3.9 first ships in Flutter 3.35, so 3.35 is
the binding floor. Pin the exact build toolchain in .fvmrc (Flutter
3.44.1).
Moving to the Dart 3.9 language level switches `dart format` to the new
"tall" style and enables two new lints. This commit is the resulting
mechanical churn, isolated from any behaviour change:
- whole-tree `dart format` reformat (tall style)
- `dart fix` for unnecessary_underscores + use_null_aware_elements
No runtime behaviour change; `make test` green.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A user-initiated denial (Deny & simplify) came back as an isError
tool_result and rendered as a prominent expanded-red "Bash · error"
block — pure noise, since the user chose it. It now folds to a muted,
collapsed "denied" card.
Built as a reusable filter rather than string-matching the note: DenyTool
carries a `quiet` flag, the session collects quiet denials' tool_use_ids,
and ConversationView renders any error whose id is in that set folded +
muted. Genuine tool failures (ids not in the set) keep the expanded-red
treatment. Adding future "expected error" cases is just adding ids.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
In live stream-json sessions a sub-agent's spawning prompt is tagged
with parent_tool_use_id, not the transcript JSONL's isSidechain +
parentUuid. The parser ignored that field, so the prompt parsed as a
main-thread user turn and rendered as a blue "you" card above the
Activity Agent card instead of folding into it.
Carry parent_tool_use_id onto ConversationItem; its presence now marks
the item as a sidechain message. The sidechain fold resolves ownership
directly by tool-use id (no transcript-only uuid chain to walk), so the
prompt folds into its Agent card and the run nests under it as before.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Every tool use now renders as a ClideCollapserCard over a one-item list
(a single tool is a list of one) — no separate single-card path. The
collapser carries the echoed last line, the count, and the aggregate
status (spinner while in-flight, check/cross once resolved); the inner
content card holds the call body + folded CALL/PROMPT/RESULT segments and
its own per-item mark. Inside a run (activity/edits/agent), tools render
as the bare inner content card so collapsers don't nest.
ConversationCard gains a `margin` param so inner cards carry no stream
margin; the collapser pads its inner canvas evenly on all sides (the
inner card no longer jams under the header). Rewrote the conversation_view
tests for the new structure and added a single-tool golden.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A run of 2+ consecutive edits to the same file now folds into one ClideHolderCard
labelled '# edits' (coalesceEditRuns, run after groupConversation) instead of a
stack of cards; a different file or an interleaving step splits the run. Every
edit stays reachable on expand.
The holder gained an optional aggregate status. New owned primitives: ClideSpinner
(the logo mark, monochrome, 3D Y-axis rotation, reduced-motion-aware) and
ClideStatusIndicator (running→spinner / success→check / error→cross, with an
AnimatedSwitcher seam for a richer transition later — kept self-contained, not
built on ConversationCard's mark). The activity card shares the same indicator.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A pasted-image @<path> token now renders as an inline, keyboard-activatable
thumbnail in the Claude conversation that opens the full image in the lightbox;
the composer's attachment chips use the same (larger, 44px) thumbnail. New
ImageThumbnail + openImageLightbox in the Claude layer; ClideMarkdown gains an
onImageToken builder seam (mirroring onRecordTap) that drops a WidgetSpan into
the text flow — it owns no Image.file/lightbox, staying generic. Missing files
degrade to a placeholder; render-only (the sent text + copyText are unchanged).
Resolves the conflicting T-236 (inline thumbnail) / T-254 (image card) designs
into the hybrid the user chose; recorded as D-89.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The conversation ListView.builder built its items (_ConversationTurn,
_ActivityCard) with no keys, so Flutter matched the stateful subtrees inside
them (ConversationCard collapse/hover/focus; ClideHolderCard expand state) to
widgets by POSITION. The visible list reshapes exactly when a tool result
lands — T-262 folds a success result into its call card and suppresses the
standalone result, errors append a sticky card, clusters re-fold — so after a
read/write completed, a card's collapse/hover state (or a cluster's identity)
could reattach to the wrong card.
Give each list item a stable ValueKey from its identity: sticky item by
item.uuid, folded cluster by its first item's uuid (namespaced turn./cluster./
run./step. so the four call sites can't collide), plus super.key on the
_ConversationTurn/_ActivityCard constructors.
Tests: unfolded cards carry per-item keys; a folded cluster carries its
first-item key.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Linkify bare ticket/governance refs (T-281, D-77, Q-5, R-2) in rendered
conversation messages so clicking one opens the record in its context-pane
reader — T- in the tickets reader, D/Q/R in the decisions reader — reusing
the existing `selection` MessageBus addressing (the same path clide ui open
and the panels use; D-6 parity already satisfied by `clide ui open`).
ClideMarkdown now linkifies bare refs in running text (paragraphs, lists,
headings, bold/italic), not just record-shaped markdown links. Matching is
word-boundary anchored so "T-shirt" (no digits) and "PT-281" (mid-word) stay
literal; `code` spans and `pre` blocks render verbatim and never reach the
linkifier, so refs inside code stay plain. The clickable span is shared
between bare refs and record-shaped links so both look and behave alike.
Tests: ClideMarkdown linkifier cases (bare T/D/Q/R tap fires onRecordTap,
T-shirt + inline-code refs stay plain, no-callback stays plain); conversation
view integration (clicking a bare ref publishes the reader-open selection to
the tickets/decisions reader).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A sub-agent's prose rendered as "claude" and its thinking as "thinking",
identical to the main assistant, because the label logic ignored
isSidechain — presenting sub-agent output as if the main Claude said it.
Now a sidechain AssistantTextMessage is labelled "agent" with a muted
stripe (never the coral claudeAccent brand), and sidechain thinking is
"agent thinking". Main-thread items are unchanged.
Tests: sidechain prose/thinking relabel, main-thread unchanged; golden
contrasting the muted agent stripe with the coral claude stripe.
This completes epic T-267 (conversation rendering streamlining): fold
success result (T-262), fold agent prompt (T-263), nest the agent run
(T-264), the shared holder primitive (T-266), and this attribution fix.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A sub-agent's sidechain run used to spill loose into the main chain,
indistinguishable from main-thread items. Now:
- _sidechainFold routes every sidechain item to its owning Agent/Task
tool-use by walking the parentUuid chain up to the Agent message it
branches off (nearest-preceding Agent as fallback) — correct even for
parallel agents.
- The run (prose / thinking / tool cards) nests in an "agent run"
ClideHolderCard UNDER the Agent card, suppressed from the top level. The
prompt still folds into the call (T-263); a successful sidechain tool
result folds into its own tool card inside the run, so it isn't a
separate step.
- When a run is shown, the Agent card's returned-result segment is dropped
(it duplicates the run's final output, note E) — but kept when no run
was captured, so output is never lost.
Tests: run nesting, returned-result dedup, parallel-run attachment (would
fail under nearest-preceding), and folded-result-not-double-counted.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A sidechain sub-agent prompt was rendered with the blue "you" label,
falsely implying the user typed it. Now:
- transcript_reader parses parentUuid (was dropped) onto every
ConversationItem.
- conversation_view resolves each sidechain prompt to its spawning
Agent/Task card via parentUuid (nearest-preceding Agent as fallback),
folds the prompt into that card as a collapsed "prompt" segment, and
suppresses the standalone block. Layered order when expanded:
call → prompt → result (note E).
- A sidechain UserMessage never gets the "you" treatment: folded into its
card, or — when orphaned — rendered as a muted standalone "agent prompt".
Tests: parentUuid parse, fold + suppression, parallel-agent attachment
(would fail under a nearest-preceding heuristic), orphan relabel, and a
golden for the Agent card's call → prompt → result layering.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A successful tool call now renders as a single merged card instead of a
separate call card + result card:
- ConversationCard gains a trailing header status slot (success check /
error cross, semantics-labelled) and optional labelled extra body
segments (P1 + P2 — reused by T-263's Agent card).
- conversation_view builds a reverse toolUseId→result lookup, folds a
successful result into its call card as a colorized ClideCodeBlock
(Read → file grammar, Bash → shell, else text), and suppresses the
standalone success result in _visibleItems. Errors stay a separate
prominent red card, now with a matching red header mark (note C).
- Suppression happens before grouping, so a merged success result is no
longer a separate activity-cluster step (note D).
Tests: card status-mark + segment coverage, merged-fold + language
inference + in-flight/orphan/error cases, cluster non-double-count, and a
golden for the merged card (collapsed + expanded + error).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Drives an image inline into the Claude conversation log over the same
bus-publish path as ui.toast/ui.open, keeping the dispatcher handler
Flutter-free. The card is display-only per D-78; the verb registers a
CommandSchema so it surfaces in clide capabilities for T-248 discovery.
Closes T-249.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A heavy agent turn buried user/Claude prose under a wall of tool-call/
result rows. A pure grouping pass (activity_cluster.dart) folds runs of
consecutive meta items into clusters; the conversation view renders each
cluster as one collapsible activity card — collapsed by default with a
live one-line ticker of the latest step + a step count, click/Enter to
expand the steps in order. Sticky items (user messages, Claude prose,
and FAILED results) render first-class and seal the cluster.
Fold level is switchable (FoldLevel none/tools/thinking/everything);
default L1 folds tool calls+results while keeping diffs and thinking
first-class. The grouping logic is fully unit-tested; the card is
keyboard + screen-reader accessible. Persisting the level via a user
setting + control is the tracked follow-up T-235.
Closes T-230 (under T-132).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The conversation pane now exploits the structured stream instead of
dumping tool input as JSON. ConversationController indexes tool_use by id
so a tool_result pairs back to its call and renders the Edit/Write diff or
is_error failure in place; per-tool bodies (Bash command+output, Read/Grep
file/query) reuse the shared renderers factored out of the permission
card. SessionStatus gains cost + contextWindow + rate-limit, read straight
off the init/result/rate_limit_event events, so the in-pane status line
reflects live state without the config probe.
Partial-message streaming is wired behind --include-partial-messages but
its event shape is unverified against the live binary and degrades to a
no-op if it differs — see T-184.
T-168.
Co-Authored-By: Claude <noreply@anthropic.com>
Two spot-check fixes (T-178, T-179), both grounded in a boundary test of
the stream-json wire (findings folded into the spike doc):
- Harness-injected user messages (skill loads, slash-command expansions,
system reminders) carry isSynthetic on the wire (isMeta in the
transcript). They were rendering as blue "you" cards though the user
never typed them; now UserMessage.injected flags them and the view
shows a muted, collapsed "context" card instead.
- Permission prompts now show the command/input being permitted (a
capped, scrollable code block) so you can see what you approve. Instead
of fully hiding a prompted tool-use, once resolved it collapses to a
one-line summary with a green (approved) or red (denied) border; the
session tracks per-tool_use_id outcome and the view colours it. The
result is kept.
Corrects an earlier wrong assumption: the Skill tool is auto-allowed
(no permission prompt); the inject only appears once the Skill tool is
actually invoked, which is why deny-captures missed it.
T-178, T-179, D-78.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
A permission-gated tool or AskUserQuestion already surfaces as a prompt
in the composer zone, so its raw tool-use card was redundant noise. The
session now tracks which tool_use_ids surfaced as a prompt; the
conversation view hides those tool-use cards. AskUserQuestion also hides
its result (the chosen answer is logged separately); permission-tool
results are kept — that's the useful outcome. The pane rebuilds the
view on each prompt change so the payload vanishes the moment its prompt
appears.
T-176, T-177, D-78.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Builds on the in-composer prompt surface (D-78):
- Permission prompts (T-175): Allow / Allow-and-don't-ask-again / Deny.
"Don't ask again" appears only when the request carries a
permission_suggestion and echoes it back as updatedPermissions. An
optional note rides Deny as the message, or Allow as a follow-up user
message (the protocol has no allow-with-message).
- AskUserQuestion picker (T-176): a single question renders bare; 2-4
questions step one at a time (nav shows "N · Header", ✓ when answered)
then a review/confirm screen. Each question offers an "Other" free-text
choice and a per-choice note; multi-select joins labels. A "chat
instead" escape denies the prompt so the user can type freely. On
submit the answer is echoed into the log, since the card is ephemeral.
- Collapsed tool cards (T-177): multi-line tool_use / tool_result start
collapsed behind a one-line summary; one-line output renders inline.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Replaces the bare "Waiting for Claude…" empty state with ClaudeBanner:
clide's own logo, a "Claude" label, the session role (primary /
secondary N), the workspace (home-collapsed), the tmux status line, and
a warming-up hint. ConversationView gains an optional emptyState widget;
the pane supplies the banner from data it already has.
Fully owned — no tmux capture-pane, no Anthropic artwork. The "Claude"
label uses Anthropic's published accent #d97757 nominatively; recorded
under a new trademark_notices section in assets/licenses.yaml (clide is
unaffiliated, bundles no Anthropic logo/artwork). Also fixes a stale
forkpty->pty reference in that file.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
test / unit + widget + golden + a11y (push) Failing after 26s
test / integration_test (xvfb) (push) Has been skipped
test / bundle smoke (xvfb 5s) (push) Has been skipped
test / daemon subprocess + web WASM smoke (push) Has been skipped
test / dart doc (lib API) (push) Failing after 23s
The Claude pane owned a TranscriptReader directly via the controller.
Insert a TranscriptPublisher that tails the transcript and republishes
each ConversationItem onto the kernel MessageBus; the view's controller
subscribes through ConversationController.fromBus. The subscription is
attached before the reader's first poll, so the initial tail isn't
missed on the broadcast bus.
This decouples reading from rendering: the team work (T-139/T-140) can
run one observer that publishes per-agent channels while the lead tile
and each teammate tile subscribe independently.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
test / unit + widget + golden + a11y (push) Failing after 27s
test / integration_test (xvfb) (push) Has been skipped
test / bundle smoke (xvfb 5s) (push) Has been skipped
test / daemon subprocess + web WASM smoke (push) Has been skipped
test / dart doc (lib API) (push) Failing after 27s
The native Claude pane (T-137) picks the newest transcript JSONL by
mtime — which on a live session is this multi-MB active file. On attach
it read and parsed the whole file synchronously on the UI isolate and
fired notifyListeners per item, freezing the app.
Three fixes: cap the initial read to a recent tail (256KB, injectable
for tests); run JSON parsing in a background isolate via Isolate.run;
coalesce the controller's notifications with a zero-duration Timer so a
burst collapses into one rebuild (a microtask-scheduled notify wouldn't
— stream events deliver one per microtask and the notify interleaves).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
test / unit + widget + golden + a11y (push) Failing after 27s
test / integration_test (xvfb) (push) Has been skipped
test / bundle smoke (xvfb 5s) (push) Has been skipped
test / daemon subprocess + web WASM smoke (push) Has been skipped
test / dart doc (lib API) (push) Failing after 27s
The Phase-1 wedge of epic T-132 (D-75): the Claude pane no longer
renders the PTY's TUI. It runs claude in tmux as before (so the
transcript is written) but displays the conversation as native cards
read from the transcript via TranscriptReader — user / assistant
markdown / thinking / tool-use / tool-result. The whole list sits under
a new no-Material ClideSelectionArea (SelectableRegion-based, since
Flutter's SelectionArea is Material and D-7 bans it), so text selects
and copies across cards — recovering the terminal's one real advantage.
ClaudePane drops its Terminal model and the resize-driven spawn trigger
(spawn now fires once on didChangeDependencies with a fixed tmux size,
since the TUI isn't shown); pane.output is no longer consumed. The
terminal builtin is untouched and still available as a general tool.
Input/composer is the next ticket (T-138).
Co-Authored-By: Claude <noreply@anthropic.com>
