Agents were bulk-reading all context task files (39 files) because
filenames are opaque IDs and there was no way to find phase-relevant
tasks without reading every file. Now writeInputFiles generates a
context/index.json with tasksByPhase mapping phaseId to task metadata
(file, id, name, status). Prompt updated to direct agents to read
the index first.
Agents were reading all contextFiles (38 tasks + 8 phases) upfront,
wasting massive context window. Updated INPUT_FILES and SESSION_STARTUP
prompts to clearly distinguish assignment files (read all) from context
files (read only when needed).
Add findByProjectId to InitiativeRepository using a subquery on the
initiative_projects junction table. Extend the listInitiatives tRPC
procedure to accept an optional projectId filter that composes with
the existing status filter. Add a project dropdown to the initiatives
page alongside the status filter.
Adds startNewSession() to useChatSession hook that closes the current
session without closing the panel. New Plus button in chat header
appears when a conversation exists, with shift+click to skip confirm.
Pass targetId to buildChatPrompt and add <scope> block that clearly
distinguishes primary target files from context files. Context entities
may be modified when necessary (e.g. dependency links) but the agent
is instructed to focus changes on the primary target.
Added retry:true flag to sendChatMessage input. Server skips storing
the user message when retry is set. Frontend uses a dedicated
retryLastMessage function that skips the optimistic message add.
When the chat agent crashes (e.g., expired OAuth token), display the
error message inline with a Retry button that re-sends the last user
message. Input stays enabled so users can also send a new message.
The createContext call manually enumerates every field rather than
spreading, so the new repo was silently dropped even though it existed
on the options object.
The repo was created in container.ts but silently dropped because
TrpcAdapterOptions (which derives ServerContextDeps) was missing the
field. This caused 500 errors on all chat session procedures.
Each execute agent has significant startup cost (context loading, codebase
exploration, baseline tests). The previous sizing guidance (sweet spot
<150 lines, 1-3 files, merge at <20 lines) produced tasks too small to
justify that overhead.
New guidance targets cohesive feature units: 200-500 lines across 3-6
files as the sweet spot, merge threshold at <100 lines, and explicit
instruction to bundle related changes (validation + route + tests = one
task, not three).
Instrument Serif is inherently condensed and illegible at heading sizes.
Playfair Display is wider, has proper variable weight support (400-900),
and reads well at all sizes.
Instrument Serif only ships weight 400, so font-bold/font-semibold
caused ugly browser-synthesized faux-bold. Tracking-tight compressed
it further. Removed both and bumped sizes slightly to let the display
font breathe.
The Execution tab's PipelinePhaseGroup was using sortByPriorityAndQueueTime
which ignores task dependencies entirely. Tasks now sort topologically via
topologicalSortPhases, matching the Plan tab's TaskGraph behavior.
- Add listInitiativeTaskDependencies tRPC procedure (bulk fetch)
- Fetch task deps in PipelineTab, group by phase, pass through
- Replace priority sort with topological sort in PipelinePhaseGroup
- Show "blocked by N" count on PipelineTaskCard
Introduces a chat loop where users send instructions to an agent that
applies changes (create/update/delete phases, tasks, pages) and stays
alive for follow-up messages. Includes schema + migration, repository
layer, chat prompt, file-io action field extension, output handler chat
mode, revert support for deletes, tRPC procedures, events, frontend
slide-over UI with inline changeset display and revert, and docs.
- Switch TaskSlideOver from absolute to fixed positioning so it overlays
the viewport instead of fighting with phase panel scroll
- Render task descriptions as rich HTML via markdownToSafeHtml (markdown
→ tiptap JSON → HTML roundtrip) with prose typography classes
- Move TaskSlideOver render to ExecutionProvider level in ExecutionTab
- Remove unnecessary relative overflow-hidden wrapper from PhaseDetailPanel
- Export markdownToSafeHtml from markdown-to-tiptap utility
- Add category color mapping utility (lib/category.ts)
- Enhance TaskNode to 2-line layout with category badge, priority, blocked count
- Create TaskSlideOver panel that animates in from right within phase detail
- Remove centered TaskModal/TaskDetailModal in favor of contextual slide-over
- Update PipelineTab to also use TaskSlideOver
Plus Jakarta Sans was too similar to Geist Sans — invisible difference.
Instrument Serif creates strong serif/sans contrast for headings, making
typographic hierarchy immediately obvious.
- Apply font-display (Plus Jakarta Sans) to all page h1/h2 headings
- Wire interactive Card prop on initiative cards and agent cards
- Redesign empty states with icons: agents (Users), inbox (Inbox/MessageSquare), right panels (Terminal)
- Unify initiative detail tabs to border-b-2 indicator style matching settings
- Strengthen two-panel dividers on agents and inbox pages (lg:border-r)
- Clean up filter pill badges — replace nested Badge with simpler span
- Increase PhaseAccordion spacing for better readability
- Fix error state styling in settings to use status tokens
- Increase settings/projects section spacing
Apply font-display to headings across settings layout and health page.
Replace text-destructive with text-status-error-fg for consistency with
the design system status tokens. Increase projects page section spacing
from space-y-4 to space-y-6.
- Add Plus Jakarta Sans as display font for headings
- Add subtle noise texture overlay + indigo radial gradient for depth
- New keyframe animations: glow-pulse, fade-in-up, scale-in, slide-in-right
- Card: interactive hover-lift + selected ring variants
- Button: scale micro-interactions, destructive glow, transition-all
- Header: logo upgrade with wordmark, animated nav indicator bar, glass search button, gradient shadow depth
- StatusDot: glow halos per status variant (active/success/error/warning/urgent)
- HealthDot: glow effects for connected/disconnected/reconnecting states
- Card hover-lift and status glow CSS utilities
Subtle fade-in + y-offset animations on mount for all main pages
(initiatives list, initiative detail, agents, inbox) and staggered
card animations for initiative and agent lists.
Tasks are now grouped by dependency depth using the same
groupPhasesByDependencyLevel utility. Parallel tasks are wrapped in dashed
containers, sequential layers connected by status-aware lines. Replaces
the flat TaskRow list and DependencyIndicator callout bars.
Phases are now grouped by dependency depth using groupPhasesByDependencyLevel.
Single-phase layers render as compact nodes, multi-phase layers are wrapped in
a dashed "PARALLEL" container. Connectors between layers turn green when prior
layers are all completed. Staggered entrance animation per layer.
Replace plain text dependency indicators with visual, status-aware components:
- New DependencyChip/PhaseNumberBadge components with status-colored styling
- Sidebar shows compact numbered circles for phase deps instead of text
- Detail panel uses bordered cards with phase badges and status indicators
- Task dependency callout bars with resolved/total counters
- Collapse mechanism for tasks with 3+ dependencies (+N more button)
- Full dark mode support via semantic status tokens
The auto-spawned agent on initiative creation was using discuss mode
(Q&A) when it should use refine mode (expand content). Now:
- Description seeds root page as tiptap content (split on double newlines)
- Spawns refine agent with the populated page in inputContext
- getActiveRefineAgent broadened to also surface discuss agents (for
CLI-spawned discuss agents)
- RefineAgentPanel shows mode-appropriate label for discuss vs refine
The discuss agent spawned on initiative creation received only the
initiative in its inputContext, missing the task that carries the user's
description. The agent started without knowing what to discuss.
Auto-spawned discuss/plan/refine agents were invisible because:
1. listInitiatives only filtered for mode='detail' agents
2. deriveInitiativeActivity returned 'idle' for zero phases before
checking for active agents
Broadened agent filter to all architect modes (discuss, plan, detail,
refine), moved active agent check before zero-phases early return, and
added 'discussing'/'refining' activity states with pulsing indicators.
Prompt changes in detail.ts invalidated the old cassette hashes.
Re-recorded all 4 cassettes with updated prompt content. Replay
verified passing in 12s.
Detail agents define task dependencies in YAML frontmatter but they were
silently dropped — never written to the task_dependencies table. This
caused all tasks to dispatch in parallel regardless of intended ordering,
and the frontend showed no dependency information.
- Add fileIdToDbId mapping and second-pass dependency creation in
output-handler.ts (mirrors existing phase dependency pattern)
- Add task_dependency to changeset entry entityType enum
- Add listPhaseTaskDependencies tRPC procedure for batch querying
- Wire blockedBy in PhaseDetailPanel and PhaseWithTasks from real data
- Clarify dependency semantics in detail prompt
Execution agents were spawning blind — no input files, no knowledge of
what predecessor tasks accomplished. This adds three capabilities:
1. summary column on tasks table — completeTask() reads the finishing
agent's result.message and stores it on the task record
2. dispatchNext() gathers full initiative context (initiative, phase,
sibling tasks, pages) and passes it as inputContext so agents get
.cw/input/task.md, initiative.md, phase.md, and context directories
3. context/tasks/*.md files now include the summary field in frontmatter
so dependent agents can see what prior agents accomplished
When an agent asks a question via `cw ask` targeting an idle agent,
the conversation router now auto-resumes the idle agent's session so
it can answer. Previously, questions to idle agents sat unanswered
forever because target resolution only matched running agents.
Changes:
- Add `resumeForConversation()` to AgentManager interface and implement
on MultiProviderAgentManager (mirrors resumeForCommit pattern)
- Relax createConversation target resolution: prefer running, fall back
to idle (was running-only)
- Trigger auto-resume after conversation creation for idle targets
- Add concurrency lock (conversationResumeLocks Set) to prevent
double-resume race conditions
Planning modes (plan, refine) get a minimal block with just cw ask
syntax. Execution modes get the full protocol: commands table, shell
recipe for listener lifecycle, targeting guidance, when/when-not
decision criteria, good/bad examples, and answering guidelines.
listInitiativeTasks was filtering out detail tasks server-side, so the
detailAgentByPhase mapping could never resolve agent.taskId to a phaseId.
Move the filter to client-side (displayTasks) so detail tasks are available
for agent mapping but excluded from counts and display grouping.
Thread detail agent info through PipelineGraph → PipelineStageColumn →
PipelinePhaseGroup. Phase groups now show spinner + "Detailing…" when a
detail agent is active and "Review changes" when finished with no tasks.
Add 'detailing' activity state derived from active detail agents
(mode=detail, status running/waiting_for_input). Initiative cards show
pulsing "Detailing" indicator. Phase sidebar items show spinner during
active detailing and "Review changes" when the agent finishes.
listInitiatives now returns an activity object (state, activePhase, phase
counts) derived server-side from phases, eliminating per-card listPhases
queries. Initiative cards show a StatusDot with pulse animation + label
instead of a static StatusBadge. Removed redundant View and Spawn Architect
buttons from cards. Added variant override prop to StatusDot.
tRPC subscriptions use connecting/pending/error/idle — not success.
The old code mapped pending→isConnecting and waited for success (which
never fires), causing AgentOutputViewer to permanently show "Connecting...".
Now: connecting→isConnecting, pending→isConnected, idle→disconnected.
Architect agents (discuss, plan, detail, refine) were producing generic
analysis disconnected from the actual codebase. They had full tool access
in their worktrees but were never instructed to explore the code.
- Add CODEBASE_EXPLORATION shared constant: read project docs, explore
structure, check existing patterns, use subagents for parallel exploration
- Inject into all 4 architect prompts after INPUT_FILES
- Strengthen discuss prompt: analysis method references codebase, examples
cite specific paths, definition_of_done requires codebase references
- Fix spawnArchitectDiscuss to pass full context (pages/phases/tasks) via
gatherInitiativeContext() — was only passing bare initiative metadata
- Update docs/agent.md with new tag ordering and shared block table
