{"slug":"architecture/WORKBENCH_EXPERIENCE","title":"IntegrateWise — Workbench Experience Package","section":"architecture","sectionLabel":"Architecture","status":"unclassified","type":"document","path":"architecture/WORKBENCH_EXPERIENCE.mdx","bytes":9339,"description":null,"updated":null,"source":"integratewise/Integratewise-docs-wiki","content":"\n**Status**: DRAFT — for product/experience review  \n**Source of truth**: `docs/architecture/WORKBENCH_MATRIX.md`, `docs/architecture/PLATFORM_LOCK_v1.0.md`  \n**Mandatory constraint**: One Workbench surface. No L1/L2/L4/L7 UI labels.\n\n---\n\n## 1. Core belief\n\nThe Workbench is not a dashboard. It is an **operational surface**.\n\nThe user does not switch contexts to work.\nThey stay in one place. Connected systems appear here when needed.\nAI stays in the background until it is valuable.\n\n---\n\n## 2. What the user experiences\n\n### 2.1 Layout\n\n```\n┌───────────────────────────────────────────────────────────────┐\n│ Workspace shell                                              │\n│ ┌─────────────────┐ ┌───────────────────────────────────────┐ │\n│ │ Sidebar         │ │ Content                               │ │\n│ │                 │ │                                       │ │\n│ │ [Brand]         │ │ Current view:                         │ │\n│ │                 │ │  • Home                                │ │\n│ │ --- Work ---    │ │  • Accounts / Contacts                │ │\n│ │ > Work          │ │  • Tasks / Calendar / Docs            │ │\n│ │   Accounts      │ │  • Analytics / Projects               │ │\n│ │   Contacts      │ │                                       │ │\n│ │   Deals         │ │ ┌───────────────────────────────────┐ │ │\n│ │   Tickets       │ │ │ Operational capability bar        │ │ │\n│ │                 │ │ │ [Call] [Email] [Slack] [Schedule] │ │ │\n│ │ --- Self ---    │ │ │ [Create task] [Open file] [Share] │ │ │\n│ │   Tasks         │ │ └───────────────────────────────────┘ │ │\n│ │   Calendar      │ │                                       │ │\n│ │   Notes         │ │ Detail cards:                         │ │\n│ │   Knowledge     │ │ • Summary / health / signals          │ │\n│ │                 │ │ • Recent activity timeline            │ │\n│ │ --- Platform ---│ │ • Connected entity list               │ │\n│ │ > Activation    │ │                                       │ │\n│ │   Integrations  │ │                                       │ │\n│ │   Configuration │ │                                       │ │\n│ │   Governance    │ │                                       │ │\n│ │   Continuity    │ │                                       │ │\n│ │   Network       │ │                                       │ │\n│ │                 │ │                                       │ │\n│ │ © Twin          │ │                                       │ │\n│ └─────────────────┘ └───────────────────────────────────────┘ │\n└───────────────────────────────────────────────────────────────┘\n```\n\nRules:\n\n- Sidebar shows **real control surfaces**, never architecture labels.\n- `Twin` is not a sidebar destination. It is a **capability** inside the Workbench.\n- Content area changes with the active module. Capability bar stays contextual.\n\n---\n\n### 2.2 User mental model\n\n| The user thinks...                    | The platform does...                           |\n| ------------------------------------- | ---------------------------------------------- |\n| Where is everything for this account? | One Account view, all connected context        |\n| How do I contact this person?         | `Call` / `Email` / `Message` in capability bar |\n| What is happening right now?          | Signals + timeline via Continuity              |\n| Can I delegate this?                  | `Assign your Twin` via OODA grammar            |\n| Is this safe to execute?              | `Approve Twin’s Action` via Governance         |\n| Who do I become here?                 | Role projection changes modules/capabilities   |\n\n---\n\n## 3. Module anatomy\n\n### 3.1 Home\n\nDefault landing inside `/app`.\n\nShows:\n\n- Active work summary for the selected role\n- Entities needing attention\n- Signals from Continuity\n- Suggested actions\n\nNo duplicate navigation. Home is a dashboard, not a menu.\n\n### 3.2 Work modules\n\nDepartment-specific operational views:\n\n- Accounts\n- Contacts\n- Deals / Opportunities\n- Tickets / Cases\n- Projects / Tasks\n- Meetings / Calendar\n- Documents / Knowledge\n- Analytics\n\nEach module shows the same **entity grammar**:\n\n```\nEntity header\n  • Identity fields\n  • Status / health\n  • Key numbers\n\nContext panel\n  • Connected records\n  • Activity timeline\n  • Linked entities\n\nOperational bar\n  • Contextual verbs only\n  • Disabled when no provider is connected\n  • Deep link when provider supports richer action\n```\n\n### 3.3 Self modules\n\nPersonal workspace:\n\n- Tasks\n- Calendar\n- Notes\n- Knowledge / Memory\n- Personal signals\n\nThese are personal projections of the same Spine data. Not a separate product.\n\n### 3.4 Platform modules\n\nReal control surfaces, always available:\n\n- **Activation Hub** — finish onboarding, enable intelligence\n- **Integration Manager** — connect providers, manage auth/sync\n- **Configuration Manager** — tenant schema, depth matrix, policies\n- **Network Graphs** — Spine relationships\n- **Governance** — approvals, audit, policy gates\n- **Continuity** — memory, context timeline\n\nNo “settings.” Everything has a name.\n\n---\n\n## 4. Operational capability grammar\n\nThe capability bar is **the same grammar everywhere**.\nOnly the available verbs change by entity type and connected providers.\n\n### 4.1 Universal verbs\n\n| Verb         | Meaning                  | Provider targets               |\n| ------------ | ------------------------ | ------------------------------ |\n| **View**     | Open canonical entity    | CRM, GitHub, Notion, Jira      |\n| **Open**     | Open in native app       | Salesforce, HubSpot, Linear    |\n| **Create**   | Create linked record     | Any connected provider         |\n| **Message**  | Send message             | Slack, Teams, Gmail, Outlook   |\n| **Call**     | Place call / open dialer | Zoom, Teams, Airwatch          |\n| **Email**    | Compose email            | Gmail, Outlook                 |\n| **Schedule** | Create meeting           | Calendar, Calendly, Zoom       |\n| **Share**    | Share record / link      | Any provider with share action |\n| **Assign**   | Delegate via Twin        | Twin runtime                   |\n| **Approve**  | Execute approved action  | Governance + Hermes            |\n\n### 4.2 Rules\n\n1. **Max 6 visible verbs** per context. Rest hidden under `More`.\n2. **Disabled state only** when no provider supports the verb. Never hide capability entirely.\n3. **Deep link always preferred** over embedding. Workbench opens native surface when action is non-trivial.\n4. **Twin verbs** (`Assign`, `Approve`) are governed. They route through the OODA grammar, not direct execution.\n\n---\n\n## 5. Twin placement\n\nTwin is **not**:\n\n- a left sidebar icon\n- a right panel\n- a chat widget\n- a button in the capability bar\n\nTwin is:\n\n- ambient signal in the header when there is material output\n- contextual suggestion inside the capability bar when user is stuck\n- surfaced via `Ask your Twin` / `Assign your Twin` buttons\n\nDefault silence law applies.\n\n---\n\n## 6. Experience quality rules\n\n1. **One mental model**: “What am I trying to do?”\n2. **No app switching penalty**: Workbench opens native app only when needed, with return path.\n3. **No duplicate data entry**: Spine is the source of truth. Connected app edits sync back.\n4. **No feature discovery tax**: Platform modules are real controls, not settings.\n5. **No dead screens**: Every view has actions or explicitly explains why actions are unavailable.\n\n---\n\n## 7. Minimum lovable Workbench (v0)\n\nDeliver exactly these views first:\n\n1. Home — work summary\n2. Accounts — operational surface\n3. Contacts — operational surface\n4. Tasks — self operational surface\n5. Calendar — self operational surface\n6. Knowledge — personal + team artifacts\n7. Integration Manager — connect providers\n8. Activation Hub — onboarding completion\n9. Governance — approvals\n10. Continuity — memory\n\nEverything else is Phase 1+.\n\n---\n\n## 8. Non-negotiables\n\n- One Workbench. No L1/L2/L4/L7 in the UI.\n- Sidebar shows Platform modules and Work modules. Nothing else.\n- Capability bar is contextual, not global.\n- Twin is silent by default.\n- AI is infrastructure. User never sees model selection, providers, tokens, or config unless explicitly needed.\n\n---\n\n**Tagline**:  \n**Truth you own. AI you rent. Approval in between.**\n"}