USE CASES

See how GhostCrab MCP
works in real projects.

GhostCrab MCP is not tied to one industry or one narrow workflow.
It becomes useful whenever your agent needs to work inside a domain that has structure:
entities, stages, dependencies, constraints, blockers, approvals, and evolving state.
This page shows what that looks like in practice.

Not just what GhostCrab MCP stores. Not just what the architecture does. But how GhostCrab MCP and the agent behave together.

INSTALL ON GITHUB READ THE ARCHITECTURE

Hello, I'm GhostCrab — your favorite MCP server.

HOW EVERY USE CASE WORKS

Every GhostCrab MCP use case follows the same pattern.

The project

What the user wants to build.

What GhostCrab MCP asks first

The setup questions asked by the MCP.

What gets modeled

Entities, phases, statuses, dependencies, constraints, and rules.

How the agent navigates

What it filters, what it follows, what it detects as missing.

What the user gets back

A useful response, next steps, blockers, and follow-up questions.

GhostCrab MCP does not just "store structured data." It helps your agent build a model, move through it, and guide work over time.

USE CASE INDEX

Onboarding

Model people, tasks, approvals, stages, blockers, and country-specific rules.

CRM Pipeline

Model companies, contacts, opportunities, objections, approvals, and next actions.

Compliance & Legal

Model obligations, prerequisites, validation steps, escalation paths, and missing requirements.

Project & Kanban

Model tasks, owners, phases, blockers, milestones, and dependencies.

Personal & Family

Model deadlines, documents, responsibilities, phases, and shared tasks.

Enterprise Command Layer

A CEO-level assistant working across strategic, financial, operational, and subsidiary-level data.

ERP & CRM Migration

Model source entities, mappings, dependencies, validation steps, and reporting goals.

01 / ONBOARDING

Onboarding

Coordinate people, equipment, compliance steps, access, timing, and blockers across a process with real dependencies.

The project

A company wants an agent that can help manage employee onboarding across teams, countries, and approval flows. The problem is not just tracking tasks. It is coordinating people, equipment, compliance steps, access, timing, and blockers across a process with real dependencies.

GhostCrab MCP asks first

→ Who are the main actors?
→ What are the onboarding stages?
→ What statuses matter?
→ What tasks depend on other tasks?
→ What changes by country, team, or role?
→ What must be approved before access is granted?
→ What counts as blocked, waiting, or complete?

Entities

employeemanagerteamequipmentaccounttrainingofficecountrypolicy

Phases

pre-hirecontract signedIT setupcompliancefirst dayprobation

Statuses

not startedin progresswaitingblockedcompleted

Dependencies

laptop required before first dayemail account required before tool accesspayroll registration required before activationcompliance training required before production access

Constraints

different rules by countrydifferent steps by teamdifferent approvals by role

How the agent navigates

→ filter onboarding cases by country, team, owner, phase, or status
→ follow dependencies between tasks and approvals
→ detect missing information
→ identify what is blocked and why
→ distinguish between waiting, blocked, and incomplete
→ ask follow-up questions when the model is not complete enough

Example

"Which new hires are blocked this week, and what should we do next?"

→ the blocked onboarding cases
→ the reason each case is blocked
→ the next best action
→ the person who needs to act
→ the missing information still required

02 / CRM PIPELINE

CRM Pipeline

Follow complex deals over time — stakeholders, objections, approvals, documents, and dependencies that evolve.

The project

A sales team wants an agent that can help follow complex deals over time. Opportunities are not just notes. They involve stakeholders, objections, approvals, documents, next steps, and dependencies that evolve over time.

GhostCrab MCP asks first

→ What are the main entities?
→ What are the pipeline stages?
→ What statuses matter?
→ What events move a deal forward?
→ What blocks a deal?
→ Which approvals are required?
→ What differs for SMB, mid-market, and enterprise deals?

Entities

companycontactopportunitymeetingproposaltaskobjectionapprovalcontract

Phases

leadqualifieddiscoveryproposalnegotiationclosed

Statuses

activestalledblockedwonlost

Dependencies

discovery before proposallegal review before signaturesecurity review before procurementpricing approval before final offer

Constraints

enterprise process rulesregional contract rulesinternal approval thresholds

How the agent navigates

→ filter deals by stage, account type, owner, or risk
→ follow dependencies between meetings, proposals, reviews, and approvals
→ detect missing stakeholders or missing documents
→ identify why a deal is stalled
→ surface the next action with the highest leverage

Example

"What is blocking this deal, and what should I do before Friday?"

→ the current blockers
→ what is missing
→ who has not responded
→ which approvals are still pending
→ the next actions in order

03 / COMPLIANCE & LEGAL WORKFLOWS

Compliance & Legal Workflows

Navigate legal or compliance processes with obligations, prerequisites, validations, escalation paths, and exceptions.

The project

A team wants an agent that can help navigate legal or compliance processes. The challenge is not just storing documents. It is understanding which rules apply, what must be done first, and what is still missing before something can move forward.

GhostCrab MCP asks first

→ What are the relevant regulations or policy sources?
→ What are the main obligations?
→ Which obligations depend on which clauses or articles?
→ What validations are required?
→ What requires human review?
→ What differs by geography, business unit, or case type?
→ What counts as complete, valid, or escalated?

Entities

regulationobligationclausepolicycontrolreviewapprovalsubmissionexception

Phases

intakeanalysisvalidationapprovalsubmissionaudit

Statuses

draftunder reviewwaitingblockedapprovedrejected

Dependencies

obligation depends on source articlesubmission requires reviewreview requires supporting evidenceescalation required for exceptions

Constraints

country-specific lawinternal policy rulesrole-based approvalsauditability requirements

How the agent navigates

→ filter obligations by region, policy, team, or review state
→ follow directional links between rules, obligations, evidence, and validations
→ detect missing requirements
→ identify contradictions or escalation points
→ ask follow-up questions when supporting data is missing

Example

"What is still missing before we can submit?"

→ the missing documents or validations
→ the blocking requirement
→ the next review step
→ the escalation path if something cannot be resolved automatically
→ the exact follow-up questions needed to complete the case

04 / PROJECT & KANBAN COORDINATION

Project & Kanban Coordination

Manage long-running work across tasks, owners, priorities, blockers, dependencies, and milestones — not just a to-do list.

The project

A team wants an agent that can help manage long-running work across tasks, owners, priorities, blockers, dependencies, and milestones. This is not just a to-do list. It is a project with structure, state, and real sequencing logic.

GhostCrab MCP asks first

→ What are the main work items?
→ What phases does the project go through?
→ What statuses matter?
→ What dependencies exist?
→ What counts as blocked?
→ What triggers escalation?
→ What milestone structure matters?

Entities

taskownermilestoneprojectteamdeliverabledependencyblocker

Phases

planningactive workreviewreleasefollow-up

Statuses

not startedin progresswaitingblockeddone

Dependencies

task A requires task Bmilestone depends on deliverablerelease requires QA approvalescalation depends on deadline risk

Constraints

ownership rulesrelease rulesresource availabilitysequence requirements

How the agent navigates

→ filter by project, owner, status, milestone, or risk
→ follow dependencies between tasks and phases
→ detect blocker chains
→ identify missing ownership or undefined next steps
→ surface priority sequences rather than flat lists

Example

"What is putting this release at risk, and what should we fix first?"

→ the blocking tasks
→ the tasks with hidden dependency risk
→ the missing owner assignments
→ the highest-priority next actions
→ the follow-up questions that would reduce uncertainty

05 / PERSONAL & FAMILY ORGANIZATION

Personal & Family Organization

Manage complex real-life coordination over time — relocation, enrollment, shared deadlines, paperwork, household logistics.

The project

A person or family wants an agent that can help manage complex real-life coordination over time. This could include relocation, school enrollment, shared deadlines, travel planning, paperwork, health administration, or household logistics. The problem is not technical. It is structural.

GhostCrab MCP asks first

→ Who are the people involved?
→ What are the main phases?
→ What deadlines matter?
→ What documents are required?
→ What depends on what?
→ What differs by country, school, provider, or case?
→ What is urgent, blocked, or waiting on someone else?

Entities

persondocumentschoolproviderdeadlineappointmenttriptask

Phases

preparationapplicationconfirmationtravelfollow-up

Statuses

pendingin progresswaitingblockedcompleted

Dependencies

application requires documentsmove depends on contracttravel depends on approvalenrollment depends on prior confirmation

Constraints

regional rulesfamily timingdocument validityshared responsibilities

How the agent navigates

→ filter by family member, deadline, provider, or urgency
→ follow dependencies between appointments, documents, and milestones
→ detect what is missing
→ track what is waiting on someone else
→ keep the whole process coherent over time

Example

"What still needs to happen before the school enrollment deadline?"

→ the missing documents
→ the next required appointments
→ the blockers still unresolved
→ the order of next actions
→ the follow-up questions needed to complete the process

GhostCrab MCP is not only for enterprise teams. It is useful whenever life itself becomes a structured project.

06 / ENTERPRISE COMMAND LAYER

Enterprise Command Layer

One assistant agent — or many specialized agents — with structured access to the operating reality of a complex organization.

The project

A top CEO wants one assistant agent to navigate the strategic reality of a large company and its subsidiaries.
Not just one business unit. Not just one dashboard.
But a structured view across financial performance, strategic initiatives, acquisitions, R&D programs,
internal operations, legal risks, subsidiary-level data, and cross-functional dependencies.

One organization may require several structured models — one for finance, one for strategic projects, one for M&A, one for legal and risk, one for subsidiary operations.

Deployment patterns

One assistant agent across multiple models

A leadership-facing assistant can access several structured domains and present a coherent high-level view.

Several specialized agents with scoped access

Each specialized agent works deeply inside its own model while still contributing to a broader strategic system.

GhostCrab MCP asks first

→ What are the major entities across the organization?
→ What strategic programs matter?
→ What are the reporting dimensions?
→ What relationships exist between parent company, subsidiaries, projects, and business units?
→ What risks, approvals, milestones, and dependencies need to be tracked?
→ Which models should stay separate?
→ Which models should be accessible through one assistant?

Entities

companysubsidiaryexecutivestrategic initiativeacquisition targetR&D programKPIreportfinancial eventlegal riskbusiness unit

Phases

planningactive executionreviewdecisionescalationclosure

Statuses

on trackat riskblockedunder reviewdelayedcompleted

Dependencies

initiative depends on budget approvalacquisition depends on due diligenceR&D milestone depends on regulatory clearancesubsidiary rollout depends on central supportboard decision depends on reporting completeness

Constraints

confidentiality boundarieslegal approval rulesreporting cadencesubsidiary separation rulesaccess scope by agent or team

How the agent navigates

→ filter the relevant slice of the organization
→ switch across several structured models
→ follow dependencies across projects, entities, and decisions
→ identify missing information before a meeting
→ surface risk concentration across subsidiaries or initiatives
→ answer strategic questions with structured grounding

Example questions

"What are the biggest strategic blockers across the company this quarter?"

"What should I review before the board meeting?"

"Which subsidiaries are most exposed if this acquisition slips?"

07 / ERP & CRM DATA MIGRATION

ERP & CRM Data Migration

Inspect, map, supervise, and finalize a migration — with structured visibility beyond the limits of the original SaaS tool.

The project

A company wants to migrate data from an ERP or CRM into a new system, a custom data model, or a new internal operating layer.
The challenge is not only moving records from one place to another.
It is understanding what the source system really contains, how entities relate, which fields map cleanly,
which fields are inconsistent, what is missing, and how to create better interfaces and dashboards
once the data is no longer trapped inside a closed SaaS product.
GhostCrab MCP becomes useful as a structured supervision and modeling layer — it does not have to perform the migration itself.

GhostCrab MCP asks first

→ What is the source system?
→ What is the target system or target model?
→ What are the main entities?
→ Which fields are critical?
→ Which relationships must be preserved?
→ Which records are authoritative?
→ What data quality problems already exist?
→ What should be normalized during migration?
→ What dashboards or interfaces are missing in the current SaaS tool?
→ What business views should become possible after migration?

Entities

companycontactopportunityinvoiceproductordercontractsupport ticketaccountcustom objectsource fieldtarget fieldmigration batch

Phases

discoveryschema reviewmappingdata cleanuptest migrationvalidationcutoverpost-migration review

Statuses

not reviewedmappedneeds normalizationinconsistentblockedvalidatedmigrated

Dependencies

target object depends on source object mappingdashboard depends on normalized dimensionscutover depends on validation completionbusiness rule depends on correct field transformationdownstream reporting depends on entity reconciliation

Constraints

source-system limitstarget schema rulesaudit requirementsbusiness continuity requirementspartial access to source dataclosed SaaS reporting limitations

How the agent navigates

→ The agent uses the ERP or CRM CLI, API, or MCP tools to inspect the source system while using GhostCrab MCP to organize the migration as a structured project. It can inspect source objects and fields, compare source structure with the target model, identify missing mappings, detect inconsistent field usage, follow dependencies between entities, flag risky transformations, track migration readiness by object or business unit, and surface the open questions that still need human validation.

Example questions

"What is still unclear before we migrate the accounts and opportunities?"

"Which field mappings are still risky?"

"What new dashboards can we build once this data is normalized?"

HOW TO THINK ABOUT THESE EXAMPLES

Every domain shares the same shape.

The point of these use cases is not that GhostCrab MCP belongs to HR, sales, compliance,
operations, family planning, or executive strategy.
The point is that all of these domains share the same shape.

entities phases statuses dependencies constraints blockers next actions

Whenever that shape exists, GhostCrab MCP becomes useful.

FAQ

What is a "project" on this page? +

A project is any domain of work that has structure over time — onboarding, a legal process, a deal, a relocation, a family workflow, a strategic initiative. If it has entities, stages, dependencies, and changing state, it behaves like a project.

What is a "model"? +

A model is the structured representation of the project — entities, phases, statuses, dependencies, rules, and constraints. A model is what makes the project navigable for the agent.

What is an "entity"? +

An entity is a thing the project is made of — a person, a task, a document, a company, a deal, a policy, a school, a milestone, a subsidiary. Entities are the building blocks of the model.

What is a "dependency"? +

A dependency is a directional relationship where one thing relies on another — a laptop must arrive before day one, legal review must happen before signature, a report must be complete before a board decision. Dependencies are what give the project its real shape.

What is a "constraint"? +

A constraint is a rule or condition that limits what can happen — country-specific regulation, approval thresholds, confidentiality rules, document validity periods, role-specific procedures. Constraints keep the model realistic.

What does GhostCrab MCP do in these examples? +

GhostCrab MCP helps define and expose the structure. It gives the agent a navigable working surface — project state, structured records, relationships, procedures, constraints, ready-to-use context.

What does the agent do? +

The agent remains the reasoning layer. It asks questions, interprets intent, follows the model, detects what is missing, explains blockers, proposes next steps, and keeps the interaction useful. GhostCrab MCP gives the structure. The agent does the thinking.

Do I need one model for everything? +

Not always. Sometimes one model is enough.
In more complex setups, it makes more sense to create several models and let one assistant or several specialized agents access them in the right way.
That is especially important in larger organizations where finance, legal, strategy, operations, and subsidiary-level work should not all live in one flat structure.

Is GhostCrab MCP only for enterprise? +

No. It works for enterprise settings, but also for personal and family use, small teams, solo operators, consultants, researchers, and anyone who wants their agent to work inside a structured domain.

Which version of MindBrain should I use? +

GhostCrab works with MindBrain Personal (SQLite) for individual use, small teams, and personal or family projects. For enterprise deployments — large teams, high volume, high availability — MindBrain Professional (PostgreSQL) offers the performance and robustness you need.

GhostCrab MCP becomes useful whenever your agent needs

to work inside a structured world.

That could be a deal, a project, a procedure, a family process,
or the strategic operating model of an entire company.
The shape changes. The value stays the same.

GhostCrab MCP helps the agent ask better questions, navigate better structure, and return better next actions.

INSTALL ON GITHUB READ THE ARCHITECTURE

Hello, I'm GhostCrab — your favorite MCP server.