You should not have to remember what is buried inside an account.
A mature Votel workspace can hold fields, pipelines, forms, agents, workflows, calendars, email designs, websites, integrations, users, and reusable values. Without documentation, onboarding, troubleshooting, handoffs, and audits turn into hours of manual clicking.
Knowledge lives in someone’s head
- Assets are discovered one screen at a time
- Dependencies are easy to miss
- Teams change things before understanding them
- “Nothing found” can be mistaken for “nothing exists”
Evidence lives in organized files
- Each accessible category has an inventory
- Important relationships and IDs are recorded
- Coverage gaps are named instead of hidden
- The next operator starts with context
One account. Twelve useful documents.
The assistant organizes the findings into a predictable pack instead of dumping hundreds of tool responses into one impossible file.
- 00Workspace OverviewPurpose, account identity, totals, major findingsSTART HERE
- 01CRM and PipelinesContacts, layouts, pipelines, stages, ownershipDOCUMENTED
- 02Fields, Tags and VariablesReusable data structure and coverage notesREVIEW
- 03Forms and WorkflowsInputs, logic, status, connectionsDOCUMENTED
- 04Agents and KnowledgeAgent configuration, purpose, sources, readinessDOCUMENTED
- 05Email and MessagingTemplates, designs, channels, configurationDOCUMENTED
- 06CalendarsBookables, staff, scheduling structureDOCUMENTED
- 07Websites and FunnelsSites, files, funnels, status, purposeDOCUMENTED
- 08Settings and IntegrationsTeams, phone, SaaS and integration contextREVIEW
- 09Asset Coverage and GapsWhat was reached, what was partial, what needs a humanREQUIRED
- 10Usage, Relationships and OrphansReferences, activity evidence, apparently unused assetsREQUIRED
- 11Workflow Enrollment and Migration ReadinessCurrent position evidence and safe re-entry planREQUIRED
Why the coverage file matters: a trustworthy document says what it could not prove. Missing access should read as unknown, never silently as empty.
Connect once, confirm twice.
Connect your preferred AI assistant
Use Claude, ChatGPT, Cursor, Claude Code, or another compatible MCP client. Follow our beginner Votel MCP connection guide.
Open the correct Votel workspace
Confirm the agency or sub-account name before the assistant reads anything. Document one authenticated workspace at a time.
Run a harmless connection test
Ask: “Tell me which Votel workspace is connected. Do not create, edit, publish, send, or delete anything.”
Choose where the blueprint should live
The prompt below asks Votel to save Markdown files in its Library. You can also ask your client to return the documents in chat or save them in your local project.
The master account-documentation prompt
This prompt deliberately separates discovery, detail reads, pagination, writing, and coverage reporting. It also forbids account changes.
Perform a read-only documentation audit of this entire Votel workspace. Inventory every accessible asset, including contact fields, contact layout, pipelines, stages, tags, global variables, email templates, forms and form elements, workflows, workflow runs and enrollments, AI agents, knowledge sources, calendars, websites, funnels, dashboards, integrations, phone numbers, users, roles, tracking campaigns, brand kits, scheduled tasks, and Library files. For every category: 1. Use its list or context tool first. 2. Follow pagination until no records remain. 3. Read each individual asset in full when a detail tool exists. 4. Record names, IDs, status, purpose, important configuration, and known relationships. 5. Never treat an unavailable tool, failed read, or permission problem as an empty category. Build a usage and relationship map. For each tag, custom field, email template, form, workflow, and reusable value, identify every accessible asset that references it. Where activity or enrollment data is available, report last activity, run counts, and which contacts are currently enrolled or recently completed a run. Classify each asset as actively used, configured but not recently observed, apparently unreferenced, or unknown because the required evidence is unavailable. For workflows, document triggers, nodes, branches, waits, messages, tags added or removed, fields read or written, templates used, calls to other workflows, and any available enrollment or run-state evidence. Do not call an asset “unused” merely because no reference was found; explain which sources were checked and which were unavailable. Do not create, edit, publish, send, activate, delete, purchase, release, invite, or overwrite anything. Do not expose credentials, tokens, secrets, private keys, or full sensitive contact data. Save the results in the Votel Library as: - 00-Workspace-Overview.md - 01-CRM-and-Pipelines.md - 02-Fields-Tags-and-Variables.md - 03-Forms-and-Workflows.md - 04-Agents-and-Knowledge.md - 05-Email-and-Messaging.md - 06-Calendars.md - 07-Websites-and-Funnels.md - 08-Settings-and-Integrations.md - 09-Asset-Coverage-and-Gaps.md - 10-Usage-Relationships-and-Orphans.md - 11-Workflow-Enrollment-and-Migration-Readiness.md For each section, cite the MCP tools used and the date checked. In the coverage report, classify each asset type as: - Fully documented - Inventory only - Inaccessible or permission-limited - Unsupported by the currently available MCP tools - Not found after a successful complete read Before reading the account, show me the documentation plan, the categories you can access, and any assumptions. Wait for my approval before continuing.
What the assistant should actually do
Discover
Read context and list tools to learn what exists and which parts of the workspace are accessible.
list_contact_fields · list_pipelines · list_forms ·
list_ai_agents
Read deeply
Open individual forms, agents, calendars, sites, dashboards, and designs where a detail tool is available.
get_form · get_ai_agent · get_bookable · get_website ·
get_dashboard
Map relationships
Record IDs, status, fields, sources, owners, dependencies, and anything another asset appears to reference.
get_contact_layout · read_agent_knowledge ·
get_workflow_status
Write the evidence
Save concise Markdown documents that a person can search, scan, hand off, and update.
save_library_file · read_library_file ·
list_library_files
Report uncertainty
Name incomplete coverage, permission failures, categories with inventory but no detail tool, and items requiring manual verification.
run_account_audit · verify_artifact
Show what is used, unused, and connected.
A list of names is only the beginning. The useful layer is a relationship graph: which workflows read a field, which actions add a tag, which templates send from which sequence, and whether there is evidence that those assets have run.
Which tags am I actually using?
Cross-check contacts carrying the tag, workflow triggers and actions, forms, agents, segments, and any other accessible references.
Unused only when both references and observed assignments are absent.Which custom fields are collecting dust?
Check contact values, layouts, forms, workflow conditions, Ask & Save nodes, pipeline cards, segments, and searches.
A blank field may still be structurally important.What is this workflow connected to?
Map triggers, branches, tag actions, field reads and writes, messages, calls to other workflows, and any run or enrollment activity.
Status, references, and actual activity are three different facts.Which templates are not being used?
Look for references from workflows, campaigns, agents, scheduled messages, and any accessible activity logs.
No reference found is not proof unless all relevant sources were checked.Use precise language: “apparently unreferenced after checking A, B, and C” is credible. “Unused” without showing the evidence is not.
Two reports turn the account into an understandable system.
The asset report tells you what exists. The connectivity report tells you how those pieces work together. Use both: an inventory without dependencies is incomplete, while a relationship map without a complete inventory can hide important gaps.
Explore the affiliate offer and open a functioning example with Overview, Assets, Connectivity, Audit, CSV, and Excel exports.
A searchable register of every accessible asset, its current status, evidence of use, and anything that deserves cleanup or human review.
- Counts and coverage by asset category
- Name, ID, type, owner, status, and last modified date
- Last observed activity and reference count
- Active, dormant, apparently unreferenced, or unknown
- Duplicates, naming conflicts, gaps, and cleanup candidates
A relationship map showing what triggers each process, what it reads or changes, and which downstream assets depend on it.
- Source → trigger → workflow → branch → action
- Tags and fields read, written, added, or removed
- Templates, agents, forms, calendars, and integrations used
- Calls to other workflows and downstream dependencies
- Run, enrollment, and contact-position evidence when exposed
| Asset | Status | References | Last evidence | Finding |
|---|---|---|---|---|
| tag:new-leadTag · tg_1042 | Actively used | 3 workflows · 248 contacts | Today | Keep; primary nurture trigger |
| lead_sourceCustom field · cf_207 | Configured | 2 forms · 1 workflow | 12 days ago | Standardize three inconsistent values |
| tag:old-campaignTag · tg_0618 | Review candidate | No accessible references found | Not observed | Verify before archive; evidence incomplete |
| sales_regionCustom field · cf_311 | Unknown | Detail read unavailable | Unknown | Requires manual review |
The practical payoff: before deleting a tag, editing a field, or replacing an email, the connectivity report shows the likely blast radius. Every conclusion should link back to its evidence and distinguish “not found” from “proven unused.”
Can a contact resume at email 5 of a 10-email workflow?
Votel can enroll contacts in workflows and provides workflow/activity reporting, but the current public documentation does not establish that a migrated contact can be inserted directly at any arbitrary internal node. Treat exact step position as data you must capture and deliberately route—not as something a normal contact import preserves.
Capture the source state
Before migration, record the source workflow, current step, last completed action, next scheduled action, and relevant timestamps for each contact.
Store a checkpoint
Import temporary fields such as
migration_sequence, migration_step,
and migration_last_action_at.
Route deliberately
Use a decision at entry—or separate continuation workflows—to place email-5 contacts on the path that begins with email 6.
Prevent duplicates
Suppress contacts whose next message already sent, respect opt-outs, and do not restart the full sequence by default.
Test a small cohort
Verify contact state, wait timing, next message, branch logic, and workflow reporting before bulk enrollment.
What can be documented—and what needs a caveat
Contact fields and layout, pipelines, forms, email designs, AI agents and knowledge, bookables, websites and files, dashboards, and many workspace settings can be documented in depth when your role has access.
Some categories may expose names and IDs without every configuration detail. Tags, relationships inferred from references, and categories without a dedicated detail tool belong here.
Visual appearance, business intent, external system behavior, credentials, hidden permissions, and anything the tools cannot access require a human confirmation.
| Area | Example reads | Blueprint treatment |
|---|---|---|
| CRM structure | Fields, layout, pipelines, contacts | Document in depth; minimize personal contact data. |
| Reusable values | Variables, tags, field references | Document what is returned; mark inventory-only categories. |
| Forms + automation | Forms, elements, status, workflows | Record configuration and status; verify inferred links. |
| Agents + knowledge | Agent configuration, knowledge sources, audit | Document purpose and readiness; never expose secrets. |
| Sites + funnels | Website metadata, files, funnel status | Document structure; visually review the live experience. |
| External services | Integration context, phone, SaaS | Record enabled state; redact credentials and confirm externally. |
Terminology note: Votel uses “global variables” for reusable workspace values in parts of the platform. Document the names, purposes, and references that MCP returns—never their secrets.
Review the blueprint before you trust it.
Update the map without starting from zero.
Run the blueprint once for the baseline, then ask for a dated comparison after major launches, team changes, migrations, or client handoffs.
Compare the current Votel workspace with the most recent Account Blueprint in the Library. Stay read-only. Recheck every documented category, follow pagination, and create a dated change report listing new, modified, missing, inaccessible, and unchanged assets. Do not overwrite the prior blueprint until I approve the comparison.
Use this workflow with Votel’s live documentation.
Tool availability changes as Votel ships. This guide was last reviewed September 17, 2026, when the authenticated Votel tool inventory showed 318 tools across 23 categories.
All Things Votel is independent and not operated by or affiliated with Votel.ai. This is a documentation workflow, not a guarantee that every Votel asset or private setting is exposed through MCP.