AI Workspace#

Workspace Home#

Path: /workspace

The workspace home is the default page after login, showing a global overview of the current workspace.

Core areas:

1. Lifecycle stage indicator: shows the current transformation stage of the workspace (L1/L2/L3/L4/L5)
2. AI daily brief card: summary of today's automatically generated brief
3. Runtime status: AI engine status, connector health
4. Cost statistics: daily/monthly AI consumption (CNY/USD switchable)
5. Quick actions: shortcuts such as New Assessment, New Scenario, New Plan
6. Recent activity: latest cross-module operations
7. Project health: health score of active projects (0-100)

Steps:

1. Click Workspace in the sidebar to enter the home page
2. Use the workspace selector at the top to switch workspaces
3. Click items in the daily brief card to jump to the corresponding module

AI Daily Brief#

The system automatically generates a workspace brief every day based on 9 signal types:

Signal	Source	Description
scenario	Decision scenarios	New/updated business scenario recommendations
risk	Risk management	Identified risk alerts
knowledge	Knowledge center	New knowledge assets, knowledge base updates
cost	Cost statistics	Abnormal AI consumption or budget reminders
approval	Approvals	Items awaiting approval
delivery	Delivery management	Milestones, task status changes
assessment	Assessment center	Assessment progress, score changes
general	General	System-level notifications

💡 The brief is cached and auto-refreshed at midnight (Redis key workspace:daily_brief, TTL = until 00:01 next day).

My Todos#

Path: /workspace/todos

Shows all to-dos of the current user, sourced from: assessment tasks, decision approvals, planning approvals, delivery tasks, AI suggestions awaiting confirmation.

Actions:

1. View the list, filterable by type/priority
2. Click an item to jump to the corresponding module
3. Mark as done to update the status automatically

AI Suggestions#

Path: /workspace/suggestions

Shows operation suggestions automatically generated by AI:

Type	Description
wbs_generate	AI-generated WBS
task_generate	AI-generated task list
schedule_optimize	Schedule optimization
risk_detect	Automatic risk detection
report_generate	AI-generated report
delay_alert	Delay warning
resource_conflict	Resource conflict reminder
scope_change	Scope change suggestion
quality_alert	Quality warning
dependency_block	Dependency blockage reminder

Suggestion states:

Actions:

1. View the list; each suggestion carries a severity level (critical/high/medium/low/info)
2. Click Accept to execute, or Dismiss to skip
3. Dismissed suggestions can be given an expiry time

Recent Activity#

Path: /workspace/activity

A timeline of the full enterprise AI transformation, in reverse chronological order, covering the L1-L6 lifecycle:

• L1 assessment creation/status changes
• L2 scenario additions/analysis
• L3 planning approval progress
• L4 project/task/milestone/risk/iteration status changes
• L5 knowledge publishing
• L6 operations activities (AI app registration, version releases, optimization proposals, user feedback, watchdog auto-rollback, re-assessment suggestions, etc.)

Activities are displayed with localized friendly copy (following the UI language, e.g. "Task completed", "Agent copilot confidence is low (0.42)"); notification delivery records do not enter the timeline (carried by the notification bell center).

Actions:

1. Browse the list (paginated, 20 per page)
2. Click an entry to jump to the related resource (L6 app activities go directly to the app detail page)

Approval Inbox#

Path: /workspace/approvals

🔒 Owner and admin only (approver's perspective): members don't see the menu or the approval badge and don't start polling; entering the URL directly shows the "No Access" page.

Shows all items awaiting your approval.

Approval sources:

• Decision approvals (L2 scenario pass/veto)
• Planning approvals (L3 plan approval)
• Delivery approvals (task deletion, budget modification, milestone publishing)

Approval actions:

Action	Description
submit_review	Submit for review
approve	Approve
conditional_approve	Conditionally approve
request_revision	Request revision
reject	Reject

Steps:

1. View the pending list in the Approval Inbox
2. Click an item to see details
3. Choose an approval action and submit

My Channels#

Path: /workspace/channels (menu: Workspace → My Channels; visible to everyone)

The tenant member's personal channel self-management page, complementing Platform Management → Connectors (tenant-level, owner/admin):

Aspect	My Channels (this page)	Platform Management → Connectors
Ownership	Personal channels (owner_user_id = you)	Tenant-level channels (not personal) + full ledger
Visibility	Only your own personal channels (mine view)	All connectors of the tenant (with ownership tags)
Creation	Always created as personally owned	Created as tenant-level
Who can use	Everyone	owner/admin only

💡 Gateways added in Platform Management are tenant-level channels and do not appear in "My Channels" — the two pages have different scopes by design.

Personal channel management:

1. Click Add Connector to create a personal channel (types/modes/credentials identical to the platform side: ClawBot QR binding, Telegram long-polling/Webhook, etc.; credentials stored encrypted)
2. Test connection; enable/disable; edit/delete
3. Expand a row to see runtime details, Webhook URL, binding status

Personal notification preferences (per-user push subscriptions):

The bottom of the page provides a "Notification Preferences" matrix — event type × channel (in-app / email / real-time push) toggles:

• One row per event type (task changes, agent completion, approval reminders, risk alerts, @mentions, task assignments, daily brief, etc.); the three toggles independently control whether that event is pushed on each channel
• Switches save immediately (not optimistic; auto-reverts on failure); all on by default
• Preferences only affect the current user's push subscriptions, not other members