Use Topline
Use Topline: Pages and workflows
Page-by-page map of every canonical customer route, its primary workflow, access boundary, important states, recovery, and deeper guide.
Topline Pages and Workflows
This is the complete customer-facing map of Topline's canonical web pages. Use
it when you know the page you are on, when a link has moved, or when you need to
understand how one page hands work to another. Legacy /settings/*, /sources/*,
/pipelines/*, /knowledge/*, /agents*, and /workforce* links redirect to
the canonical routes below and are not separate product surfaces.
Storybook is at /storybook. Existing /storyboard, /account/storyboard,
/account/storybook, and /settings/design links redirect there while keeping
the conversation query and URL fragment.
Navigation and access
- Chat is the default signed-in surface at
/and/chat. - Apps & Workspaces at
/projectsis the library for workspaces, apps, and saved outputs. - Account at
/accountopens the first visible account section, normally Profile. Its rail groups visible sections under Personal, Workspace, Organization, and Activity. - Quick open searches pages, workspaces, apps, and chats that the current operator may access.
- A hidden destination or an Access required state means the current tenant or user lacks the capability. A copied URL never bypasses that check.
Preserve the current tenant and workspace context when diagnosing a missing page. Do not infer another customer, account, or workspace from a browser tab, old link, or remembered task.
Public and entry pages
| Page | Canonical route | Primary workflow | Important state or recovery | Guide |
|---|---|---|---|---|
| Sign in | /login | Authenticate, then return only to a validated Topline destination. | A session-loading failure is retryable; an access denial requires an administrator, not a different user's link. | Sign in |
| Fit check and signup | /fit-check, /signup | Submit the organization and workflow fit check, preserve the confirmation, and wait for Topline follow-up. | Submission creates an intake lead, not a provisioned customer account. Retry a failed submission without duplicating a successful one. | Fit check and signup |
| Documentation | /docs, /docs/:docId | Browse by product, search the curated catalog, open a stable document URL, and report helpful, unclear, or outdated content. | Clear a zero-result search or retry an index failure. Public docs never expose internal runbooks. | Docs |
| Customer appointment scheduling | /schedule/:link | Open a business-issued link, receive a one-time code at a masked account contact, verify the code, select eligible services for one visit, choose an arrival window, review the details, and confirm once. | Service details remain hidden until verification. Arrival windows are not held before confirmation. If the booking response is lost, recover the original booking before trying again; pending or partial results show a booking reference for support. Invalid, revoked, expired, or ineligible links fail closed. | Customer appointment scheduling |
| Published artifact | /p/:token | Open the approved live artifact represented by a restricted or public viewer link. | Revoked, expired, or unauthorized links fail closed and do not reveal draft content. | Artifact app and published-link access |
| Embedded artifact | /embed/:token | Render an approved live artifact inside a supported host. | Embeds inherit the published viewer policy; they are not a draft-review or editing path. | Artifact addresses, domains, and embeds |
See Use Topline: Product guide for the product model and Artifacts, Assets, and Workspaces for sharing and publishing boundaries.
Chat, apps, and workspace pages
| Page | Canonical route | Primary workflow | Important state or recovery | Guide |
|---|---|---|---|---|
| Main chat | /, /chat | Start or resume a personal or workspace conversation, attach approved files, use @ to add a teammate in an owned thread or attach exact app, artifact, and schedule context, create threads, review todos and attention items, and save durable output. A thread participant may switch between automatic replies and Mention only, where a standalone @Topline triggers the response and other messages remain shared context. | Adding a person changes the thread roster. Verify the roster or context pill; plain mention text is not success. Verify the response-mode control after a failed change. Retry a failed send from the same conversation instead of creating duplicates. | Chat |
| Apps & Workspaces | /projects | Choose All workspaces or an exact workspace, then browse and filter its apps and outputs; create a workspace, start a thread, move owned unfiled items, or open an item's details and symbol settings. | The catalog heading confirms the active workspace scope. Unfiled remains separate. View-only items cannot be moved, and symbol editing appears only with write access. Clear filters or change workspace before treating an empty result as missing data. | Workspaces |
| Knowledge Bases | /knowledge-bases, /knowledge-bases/:id | Create a distinct document collection; import HTML, files, URLs, sitemaps, or managed sources; monitor and schedule draft refreshes; restrict individual documents to selected connections; compare and test a candidate; publish; and copy its HTML or OAuth MCP endpoint. Mention a published knowledge base in Topline chat to narrow that turn, or bind it explicitly to an artifact or compatible harness. | Imports and scheduled refreshes create draft changes only. Required retrieval tests gate the exact candidate. Document policies are frozen per release. Paused, revoked, unpublished, cross-resource OAuth, or unauthorized access fails closed. | Knowledge Bases |
| Workspace home | /projects/:projectId | Review overview, instructions, documents, skills, apps, outputs, and conversations; start a thread; manage access; rename, change the workspace symbol, or archive when authorized. | A workspace owns its main chat and threads. Shared members cannot change owner-only settings or the workspace symbol. | Workspace home |
| Archived workspaces | /projects/archived | Restore an archived workspace or review permanent deletion. | Permanent deletion is blocked while dependencies or active work remain and requires typing the exact workspace name. | Archive, restore, and delete workspaces |
| Artifact detail | /artifacts/:id | Inspect the current draft or live result, versions, source, data freshness, sharing, schedules, validation, and symbol. Direct drawer links use ?drawer=details or symbol; automated tests remain in the build and validation workflow. | Draft access and live viewer access are different. Viewers cannot change the artifact symbol. A successful deploy is not live promotion. Describe problems in normal chat with the artifact attached or mentioned; default Coding agent instructions require a regression test before the fix and relevant suite run. | Artifacts |
| Assets | /assets | Organize reusable files in folders, upload, preview, rename, move, download, copy a stable reference, or delete. | Folder deletion requires an empty folder. A failed upload or move preserves the existing library state. | Assets |
| Storybook | /storybook | Describe a visual change, review the bounded draft and validation, request a correction, publish, or restore a prior reviewed version. | Keep Storybook work in its owning conversation. A Storybook story or source commit is not a published customer change. | Storybook |
| Feedback | /feedback | Submit a bug, polish request, or feature request with bounded details and optional images; follow the queued coding task. | The draft and selected images remain available after submission failure. Do not attach secrets or unnecessary PII. | Feedback |
The detailed library, artifact, asset, workspace-lifecycle, sharing, and Storybook procedures are in Artifacts, Assets, and Workspaces. Delegated execution is covered by Workforce and Storybook.
Scheduled work and secure browser pages
| Page | Canonical route | Primary workflow | Important state or recovery | Guide |
|---|---|---|---|---|
| Scheduled Work | /automations | Search schedules and reminders, create one, pause or resume, run now when supported, cancel a reminder, and inspect status. | Confirm timezone, next run, target, and recent outcome before saving or enabling. | Scheduled Work |
| Schedule detail | /schedules/:id | Review or edit schedule settings and inspect run history. | App-managed schedules may be read-only here; change their owning app manifest instead. | Schedule details and run history |
| Secure browser | /browser | Start or resume an isolated browser session, enter sensitive sign-in values yourself, take control, and end the session. | An expired or unavailable session must be reopened. Never put passwords, MFA, payment data, or recovery values in chat. | Secure browser |
See Scheduled Work and Secure Browser for complete controls, security, and recovery guidance.
Assistant pages
The Assistant hub at /account/assistant opens each tool in a drawer so the
Account context remains visible. Older direct routes such as
/account/glossary redirect to the corresponding panel.
| Panel or detail | Canonical route | Primary workflow | Important state or recovery | Guide |
|---|---|---|---|---|
| Assistant hub | /account/assistant | Choose Preferences, Company Knowledge, Glossary, Docs, Skill Library, Chat Integrations, Developer Access, or Coding Agents. | Only panels allowed by current capabilities appear. | Assistant |
| Preferences | /account/assistant?panel=preferences | Add, correct, or remove explicit personal operating guidance. A non-conflicting save becomes active immediately. | A conflict candidate remains inactive until it is reviewed and resolved. | Preferences |
| Question assistance | /account/question-assistance | Set how involved to be, the wait before automatic answers, explanation detail, question categories, memory use, and specific instructions. Task settings can override personal defaults. | Automatic answers require opt-in and a reversible technical choice. Business facts and approvals still wait for the user. Write my own answer, Why this answer?, and explicit personal/task/shared learning controls remain available. Shared corrections require review. Availability depends on release and account access. | Question assistance |
| Company Knowledge | /account/assistant?panel=knowledge | Review durable company facts and resolve conflicts before publication. | Store source-backed durable knowledge, not secrets, transient tasks, or raw customer rows. | Company Knowledge |
| Glossary | /account/assistant?panel=glossary | Search, propose, publish, edit, or retire approved business terms. | Unknown terminology remains unknown until an authorized steward resolves it. | Glossary |
| Docs | /account/assistant?panel=docs | Search the same curated documentation catalog without leaving Account. | Internal documents are visible only through authorized internal tools, not the public catalog. | Docs |
| Skill Library | /account/assistant?panel=skills, /account/skill-library/:skillId | Create or import a skill, inspect files and versions, edit customer-owned skills, and review run safeguards and allowed domains. | Every skill keeps a root SKILL.md; Topline-maintained skills are read-only. | Skill Library |
| Chat Integrations | /account/assistant?panel=chat | Configure Slack or Google Chat, reconnect Slack after a native agent upgrade, inspect captured spaces, and request a bounded historical import. | Slack Ready/Active labels do not certify native sessions; verify a real DM and the Slack app configuration. Shared Google Chat spaces require explicit platform enrollment. | Chat integrations |
| Developer Access | /account/assistant?panel=developer | Configure Build or Workspace MCP, OAuth, fallback tokens, personal Drive/Calendar access, and connected MCP tool policy. | Each product and connection has separate scopes; Check connection cannot inspect an external client's local configuration. | Developer Access |
| Coding Agents | /account/assistant?panel=coding, /account/coding-agents, /account/coding-agents/:agentId | Create or find an eligible coding agent, then inspect, configure, or remove it and review its workspace, model, credentials, and bridge readiness. | Removing an agent does not prove that related source, tasks, or deployed artifacts were removed. | Coding Agents |
See Assistant Knowledge, Knowledge Bases, Slack and Google Chat, MCP Server Reference, and Workforce and Storybook.
Data pages
| Page | Canonical route | Primary workflow | Important state or recovery | Guide |
|---|---|---|---|---|
| Data Sources | /account/sources | Find connected sources, add a source, and review setup, table freshness, sync failures, and recent activity on one page. | A filtered empty state is not an empty account. Customers can build before connecting a source. Start recovery from the latest failed sync. | Data Sources |
| Add data source | /account/sources/new | Choose a supported connector, save a credential reference, validate access, select bounded resources, and create the source. | A saved source may still need runtime setup or an initial load. Never paste credentials into documentation or chat. | Add data source |
| Connect PostgreSQL | /account/sources/new/postgresql | Validate a read-only PostgreSQL login over TLS, select tables and columns, and configure the first bounded sync. Return to the saved source to review runs or rotate credentials. | Every request requires one authenticated workspace tenant. Missing or conflicting tenant context fails closed. Credentials remain in the secret store; setup does not authorize deployment, a schema migration, or customer rollout. | Connect PostgreSQL |
| Source detail | /account/sources/:sourceId | Review connection health, manage selected tables, inspect previews and schema, run or recover syncs, and copy redacted support details. | Retry an initial load from its saved checkpoint only after its blocker is corrected. | Source detail |
| Google Sheets | /account/sources/google-sheets | Add, find, run, and recover managed Google Sheets sources. | For direct task access to an existing sheet without ingestion, use the separate sharing workflow instead. | Google Sheets |
| Offices | /account/sources/offices | Create and maintain tenant office records and additional office fields used by scheduling and reporting. | Validate addresses and time zones before relying on route or schedule results. | Offices |
| Employee Identity | /account/sources/identity | Review provider-account matches and merge evidence-backed identities. | A merge is consequential and must not be inferred from similar names alone. | Employee Identity |
| Reporting tables | /account/sources/tables | Search tables, filter by source, and identify stale or failed data. | Open the responsible source or sync before retrying. | Reporting tables |
| Table access | /account/sources/access | Account Admins review every queryable table, including tables built inside artifacts, and who can query it. | Everyone in the account can query every table today; table and column restrictions are not available yet. | Table access |
| Table detail | /account/sources/:sourceId/tables/:tableName | Inspect sample rows, columns, sensitivity, trust, relationships, unmapped fields, and sync activity; ask chat about the table. | Preview rows are bounded samples. New unmapped fields are excluded until reviewed. | Table detail |
| New table sync | /account/pipelines/new | Start from a source, select a supported missing table, choose columns, and create its sync. | If no source is selected, return to Data Sources rather than inventing a connection. | New table sync |
| Pipeline detail | /account/pipelines/:pipelineId | Redirect to the owning source and focused sync controls. | The source is the canonical control room; old standalone pipeline links are compatibility routes. | New table sync |
| Trusted Calculations | /account/trusted-calculations | Draft a business calculation from an intent, review generated logic and validation, revise, publish, or retire. | Only published, validated definitions are trusted. Retirement preserves history. | Trusted Calculations |
See Data Sources, Tables, and Pipelines, Use an existing Google Sheet, and Scheduling.
Account, access, billing, and monitoring pages
| Page | Canonical route | Primary workflow | Important state or recovery | Guide |
|---|---|---|---|---|
| Profile | /account/profile | Update the current user's display name, inspect account access, and sign out. | Identity, role, and tenant access are administrator-owned even when the display name is editable. | Profile |
| Billing | /account/billing | Review the billing period, invoice status, provider and product line items, pricing, and authorized exports; customer admins also set organization and per-person coding limits, pause new paid usage immediately, select active-admin email recipients, and inspect forecasts, task charges, and alert history. | Missing, delayed, or unreconciled provider data is unknown, never zero. Coding limits count settled and pending reservations; forecasts are directional; organization thresholds are durable and emailed once per month to selected active admins. | Billing |
| Usage | /account/usage | Review personal feature usage by month; billing viewers can switch to organization totals and people. Manage personal daily coding limits; billing managers can define tiers with daily allowances, assign people and grant unlimited user allowances within the organization cap. | Personal views are bound to the authenticated identity. Daily coding allowances reset at 00:00 UTC; pending work and monthly organization caps still apply. Unknown data remains unknown, and errors offer retry or a specific recovery path. | Usage |
| Resources | /account/resources | Review database and cache capacity, submit a justified capacity request, and follow its status and history. | Resource requests are reviewable control-plane actions, not immediate infrastructure changes. | Resources |
| Team Members | /account/members | Add or invite members, select access type and roles, manage workspace or app access, resend an invite, or archive access. | Verify one intended and one prohibited action after a grant; offboarding must also remove resource grants and owned credentials. | Team Members |
| Data access (draft) | /account/data-access | Active tenant Admins prepare exact member/team reporting-table grants with review and save. Admins have current/future all-table reporting access. | Shown only where Topline has turned on the draft. Production activation is blocked pending saved-output/delivery certification. Off-mode saves do not restrict existing queries; conflicts preserve choices for fresh review. | Data access |
| Customer scheduling | /account/scheduling | Save branding and verification delivery, review Launch prerequisites and eligibility policy, check unresolved booking requests, and create/reset or revoke customer links. | Link creation requires working delivery, a saved enabled eligibility policy, and a ready scheduling engine. Accepted delivery tests alone do not certify readiness; unresolved pending/partial bookings need reconciliation before another attempt. | Customer scheduling |
| Call review | /account/call-review | An account administrator loads a registered Five9 call, then explicitly starts xAI Grok transcription and an optional summary draft. Review the transcript and supporting evidence before using a suggested outcome. | Loading saved results does not start paid processing. Missing recordings remain incomplete, unknown costs remain unknown, and uncertain provider outcomes require review before retrying. Availability requires account setup and enablement. | Call review |
| SMS setup | /account/sms | Save company and messaging details, choose toll-free or local-number preferences, provide consent and agreement evidence, and submit for Topline-managed review. | Saving or submitting intake does not buy a number, register a sender, sign an agreement, or enable sending. Resolve requested changes and save before opening an agreement. Availability depends on release and account access. | SMS setup |
| SMS policy workspace | /account/sms/workspace | Find policy drafts after SMS agreement signing, edit saved documents, submit a snapshot for Topline review and explicitly publish approved customer-branded policies. Signers can retry incomplete workspace setup without re-signing. | Separate from general onboarding. Existing permissions and preserved edits/removals. Customer-domain hosting and legal review precede signer publication; public URLs stay stable through reviewed updates. Draft edits never change live pages. Publication is not carrier approval or sending authority. | SMS policy workspace |
| SMS Do not contact | /account/sms/dnc | Add numbers or import a reviewed one-column CSV, check a masked restriction and inspect recent activity. | Blocks Topline SMS across the customer's apps and numbers. Release requires an authorized reviewer and exact-number confirmation. No calling, external dialer or registry integration; no sending enablement. | SMS Do not contact |
| SMS setup demo | /account/sms/demo | Explore a synthetic intake and review walkthrough. | Demo changes stay in memory; they do not save a customer application or call a messaging provider. | SMS setup demo |
| SMS agreement | /account/sms/:intakeId/agreement | Review the exact terms and messaging scope prepared by Topline; the designated signed-in representative can accept when signing is enabled and retain a downloadable receipt. | Signing is disabled by default pending counsel-approved terms and operational review. Changed scope, expired or withdrawn offers require a fresh offer. Existing accepted receipts remain readable when signing is disabled. Agreement acceptance is not recipient SMS consent, carrier approval, or permission to send. | SMS agreement |
| Credentials | /account/credentials | Create a named secret reference, inspect Used in and rotation reasons, rotate the active value, make a connection-private credential shared, or delete an unused reference. | Secret values are write-only. Resolve consumers before destructive removal. | Credentials |
| Brand | /account/brand | Set the workspace name, logo, and favicon from approved files or the asset library. | A fallback remains until the custom asset saves successfully; decorative images still require an accessible workspace name. | Brand |
| Domains | /account/domains | Connect multiple company-owned subdomains, verify DNS ownership, track TLS and edge provisioning, and choose the default initial domain for new artifact app addresses. | Existing routes do not move when the default changes. Only connected domains can serve artifacts or become the default; move every artifact route before removal. Auth0 sign-in remains on Topline's stable identity domain and returns through an artifact-scoped handoff. See Company domains for artifact apps. | Domains |
| Audit Log | /account/audit-log | Filter by time, actor, action, resource, or outcome, then inspect a bounded event. | Audit metadata proves recorded platform activity, not deployment or external-provider state. | Audit Log |
| App Builder Health | /account/app-builder-health | Review sandbox, deployment, storage, source, and production-gate checks in priority order. | Fix the first blocking dependency and refresh; a healthy builder does not authorize deployment or promotion. | App Builder Health |
| Task Monitor | /account/task-monitor, /account/task-monitor/:runId | Review approvals, active work, history, plan, execution, and verification; open the owning conversation when needed. | Do not start a duplicate task for an item that only needs attention. Completion is distinct from merge, deployment, and customer rollout. | Task Monitor |
See Tenant Administration and Workforce and Storybook.
Cross-page workflows
Conversation to durable workspace output
- Start in the correct personal or workspace chat.
- Clarify the outcome, tenant, data scope, and acceptance criteria.
- Create a thread for a bounded subtopic or delegate a coding task when source changes are required.
- Save the reviewed result as an artifact or app.
- File owned output into the correct workspace in Apps & Workspaces.
- Share a draft with named collaborators or publish an approved live version; do not treat those access modes as interchangeable.
Data to recurring output
- Connect and validate the source.
- Review selected resources, reporting columns, sensitivity, and trust.
- Complete the first bounded sync and inspect the table preview.
- Create or review a trusted calculation when business logic is reusable.
- Build the app or output and label its data freshness honestly.
- Add a schedule only after a successful manual run, then monitor its history.
Report a product problem
- Open Feedback from the failing route so page context can be included.
- Describe expected behavior, actual behavior, and safe reproduction steps.
- Attach only bounded screenshots with secrets and unnecessary PII removed.
- Submit once and follow the resulting task instead of creating duplicates.
Shared states and recovery
- Loading means no conclusion can be drawn yet.
- Empty means the current unfiltered query returned no items.
- No matches means search or filters are hiding available items; clear them before changing configuration.
- Needs attention means a specific item or prerequisite requires action; open its detail rather than creating a replacement.
- Access required is an authorization boundary, not a navigation failure.
- Unavailable or failed should preserve the draft or last known safe state and offer a bounded retry.
Security, automation, and MCP boundaries
Every page remains tenant-bound and capability-checked by the API. Links, artifact tokens, workspace membership, member roles, OAuth scopes, connected MCP tool policy, and task-scoped workforce grants are separate authorities. Granting one does not imply the others.
Automations repeat only their saved target and authority. Build MCP manages artifact-build workflows; Workspace MCP exposes permission-filtered Topline context and reviewed knowledge operations. Neither product silently grants tenant administration, live promotion, cloud deployment, or customer rollout.
Documentation smoke check
For a changed page or workflow:
- confirm its canonical route and visible interface terminology;
- verify the shortest successful path with an eligible user;
- verify loading, empty, error, access-denied, and consequential confirmation states that apply;
- check the deeper guide linked from this page;
- run the documentation generation, link, search, and rendering checks; and
- verify the deployed documentation separately from source integration.
Current limitations
Available pages, connectors, models, external-chat surfaces, and controls vary by tenant capability and rollout. Some legacy URLs remain as redirects. Public artifact links expose only the approved live viewer surface, and external provider or deployment status always requires verification from its owning system.