General (identity, mode, workspaces), Providers, Skills, and Advanced
(Composio key, the operator request console, billing, migration) replace
the six-tab sprawl. Card-in-card nesting and every decorative subtitle are
gone; row actions - provider Check/Manage/Revoke, workspace Add, skill
Import, migration Report - reveal on hover or focus instead of shouting all
the time. The Composio panel loses its hardcoded org-wide toolkit grid,
which the Connectors grants model had already made dead weight, and keeps
only the API key. Fixing the flow also surfaced a real bug: provider policy
rows remounted on a snapshot-derived key, so any refresh mid-edit wiped the
operator's unsaved choices - rows now keep stable identity and edits
survive. E2e follows the new layout.
The Social surface gives way to Connectors: the searchable Composio catalog
(letter-glyph tiles - the CSP stays closed to remote images), one-click
connect that opens the authorization link in the browser and polls the
account until it turns active, and a Connected section where each toolkit
shows its status and a chip per agent that grants or revokes their access
on the spot. Route, nav label, glyph, and search targets follow the rename;
the social backend services remain for the domain while their page and e2e
steps retire in favor of connector coverage.
The office e2e now also hires two agents through the real Team flow,
returns to the floor, clicks the first open desk through the same pure
floor-plan and camera math the app renders with, and asserts the right
agent's card opens - then captures a populated-floor screenshot for visual
review. Nav locators use exact accessible names so 'Chief Executive
Officer' cards cannot shadow the Office nav button.
Canvas clicks hit-test against the exact positions the renderer draws -
frontmost person wins - and select the agent: a floating card in the app's
design language shows their name, title, department, level, live status, and
current task, with a selection ring under their feet. Escape, the close
button, or clicking empty floor dismisses it; hovering a person shows a
pointer cursor. The floor now seats only lifecycle-active agents, matching
how the backend wakes and assigns work. Covered by a hit-test unit test and
a dedicated office e2e that clicks the CEO through the real floor-plan math.
Replace the viewport-derived desk grid with a fixed-world office: a north
band of rooms (CEO office, meeting room, snack bar, restroom) with door
signage opening onto a corridor, and an open desk field that grows by rows
so any head-count fits the world instead of being cut off. Geometry lives in
a pure floor-plan module that also builds an axis-aligned lane graph with
BFS routing, unit-tested for reachability and collision invariants. A camera
transform centers or clamps the view, the first c_suite agent takes the CEO
office, name labels render the full name, and the mock TODO-board text is
gone. Office screenshots join the visual-review capture set.
The virtual office is now a dedicated page (nav: Office) with an HTML5
Canvas rendering animated pixel-art characters. Each agent is drawn as
a pixel person with deterministic skin/shirt colors, walking legs, and
a bobbing head when active. Idle agents wander randomly near their desk.
Active agents show a green dot, thinking agents show a question mark.
Departments are labeled zones with desks and monitors. Removed the old
sidebar component entirely.
E2e test updated to verify: virtual office sidebar toggle, Composio
Tools settings tab visibility, and @mention with canonical handles.
New integration test verifies that a human reply in a thread wakes
agent participants without requiring explicit @mentions.
The thread panel rendered an "Audit" feed — "Chat message sent", "Task
run completed", "Approval requested" rows with monospace timestamps —
a server log sitting inside a conversation. Attachments showed an eight-
character SHA-256 fragment next to the file type. Neither belongs in a
surface a non-technical operator reads as people talking.
Remove the audit replay from the chat thread (and its now-dead
replayEventsForThread/auditPayload plumbing) and drop the checksum from
the attachment card. The audit trail itself is unchanged — it is still
recorded and still shown on the task activity tab. The e2e thread check
now asserts the reply box, not the log.
Settings shrinks from 14 tabs to the five that serve the product:
Workspace, Billing, Providers, Skills, Migration. The deleted panels
(Members/IAM, Models, Policies, Operations, SDLC, Compliance, Profile,
Notifications, Security) were read-only governance tables and raw
audit-log dumps with no old-app counterpart. The Automation page goes
too: its rules stored empty conditions and no engine ever executed
them. Sidebar reordered to match the old app hierarchy (Home, Inbox,
Work, Approvals first). Backend services and schema are untouched.
A goal created from the Work page only wrote a goals row; nothing ever
read it, so the company never started working on it. Creating a goal now
produces an "Operationalize goal" kickoff task for the chosen owner (or
the top-level agent when unowned) and re-enables that agent's heartbeat,
so the autonomous queue worker starts the run - the same activation
contract the previous runtime had via goal_activated wakes and CEO
bootstrap. Onboarding shares the same kickoff module instead of its own
copy.
The new app reached a real LLM but never ran an agent: the kickoff task_run
sat queued forever and runTask was a single text completion with no tools.
This adds the missing runtime:
- provider-tool-turn: OpenAI Responses + Anthropic Messages tool-use turns
- agent-tools + deliverable-writer: hire_agent, create_task,
update_task_status, write_deliverable (workspace-scoped, traversal-safe)
- agent-run-loop: real tool-use loop, capped iterations
- task-queue-worker: claims queued runs in a transaction and executes them
task-execution now drives the loop; a background worker auto-runs the
bootstrap CEO run so it hires its team, delegates tasks, and writes the first
deliverable into the bound workspace. Autonomy is gated by AGENTCOMPANY_NEW_AUTORUN
(tests run with it off to stay hermetic).
Removes "Nothing pending.", "No tasks…", "Empty.", "No messages…", "No
files…", and similar status descriptions across Home/Work/Chat/Files/
Approvals/Social/Team/Settings panels — design-source never renders this
copy, and the user-frontend rule forbids status narration.
Lets the snapshot loader throw on IPC failure instead of holding a banner
string; the shell topbar handles its own crash boundary.
Fixes Team org-chart connector when only one agent is hired, switches
Settings to use one breadcrumb path, drops the unread badge until we have
read-state tracking, and snaps the Providers grid to the two-column
design spec.
- Port design tokens, sidebar, topbar, and icons to inline SVG set
- Rewrite each page (Home/Team/Work/Chat/Files/Approvals/Social/Settings)
to drive real snapshot data and match the bento, kanban, three-pane chat,
file tree + pinned cards, approval queue, social schedule + calendar,
and settings sub-nav layouts
- Split Chat and Settings into per-concern subcomponents so every file
stays under 500 lines
- Add schema v4 for pinned files and scheduled posts plus IPC + service
wiring; cover new behavior in unit tests
- Update onboarding labels and Hire form to remove ambiguity for the
expanded e2e suite (all settings tabs + 8 routes at 1280 and 980)
ZacharyZhang-NY
