What teams use the Operating MCP for

This article assumes you've already connected the Operating MCP to an LLM — Claude, ChatGPT, or Gemini. If you haven't, start with the setup guide first.

https://support.operating.app/dashboard/articles/feubglzdz0thp5ps/2817668

Once connected, your AI assistant has live read and write access to your Operating data. The examples below are meant to spark ideas, not define the limits of what's possible — every team's workflow is a little different, and the open-ended nature of a natural language interface means you can often just describe what you need and see what happens.

Note: All screenshots are taken from dummy data and may not reflect real-world scenarios. Some are cropped as the full model response was longer than what's shown. The Operating MCP is currently in beta — features and available tools may change.

From proposal to staffed project

The gap between winning a deal and actually setting up the work in Operating is one of the most common friction points for operations managers. You have a signed proposal, a scope of work, agreed dates, and a team shape — and then you spend 15–20 minutes clicking through the UI to recreate it.

With MCP, you can paste the relevant section of your SOW directly into the chat and ask the LLM to create the full project structure in Operating. It will check whether the client already exists, create the project with the right billing type, set up phases, create positions for each role mentioned, and attach the budget — all in one conversation.

This same conversation can continue into resourcing: once the project is created, you can immediately ask who's available to fill the open positions.

Staffing and capacity visibility

Staffing managers typically need to reconcile two separate views: what work is coming, and who's available to do it. An LLM with MCP access can combine them on demand.

You can ask things like:

• Who on the engineering team has capacity in August?

• Which of our active projects are understaffed for the next quarter?

• Is anyone over 100% allocated in June, and what's causing it?

• Who's been on the bench for more than two weeks?

This is also where staffing and hiring gap analysis becomes practical. You can ask the LLM to look at your confirmed project pipeline versus your current capacity — by role, seniority, or site — and surface where the gaps are across a specific timeframe.

Utilization and revenue reports

Operating has dedicated report tools that the MCP can call directly — these aren't generic data queries, they're the same underlying calculations the platform uses for its built-in reports. The LLM can pull these on demand with whatever filters and bucketing you need.

For planned reports, you can ask for utilization per person grouped by week or month, filtered by site, role, group, or even skill. You can also weight tentative work by deal win probability — so if a pipeline deal has a 60% close probability, it contributes 60% of its hours to the forecast rather than the full amount. This is useful for realistic capacity planning when you have a mix of confirmed and pipeline work.

For actuals reports, you can ask for earned revenue and tracked hours per project or per person over any date range, with breakdowns by individual contributor.

The difference between this and just "querying your data" is that these are structured financial calculations, not just raw record lookups. The LLM can combine a few calls and give you a synthesized answer that would take several report tabs and a spreadsheet to assemble manually.

Tracking time without opening the app

For consultants, the most everyday use of Operating MCP is simply logging time. If your LLM is already open — in Claude, Claude Code, a Gemini CLI session, or ChatGPT — you can log time without switching to the Operating UI.

This works especially well when you want to log several entries at once from notes at the end of the day, or when you need to batch-enter time for a team when you have the data in a table.

Invoice tracking and billing readiness

The MCP can read your invoice data across projects and clients, which is useful for the ops manager who needs to know where things stand at any given moment without opening every project individually.

You can ask:

• Which invoices are ready to send but haven't been yet?

• Are there any invoices past their due date?

• What's been invoiced to a specific client this quarter?

This is particularly useful during month-end: rather than checking project by project in the UI, you can get a consolidated view of billing status across the portfolio in one pass.

Onboarding and managing people

When a new person joins, you can add them to Operating via MCP — useful if you're setting up several people at once or want to do it from a chat interface without navigating through the settings UI.

The same applies to updating existing people: changing a role, adjusting a site assignment, or archiving someone whose contract has ended. If you're handling an employment lifecycle event, you can describe the change conversationally and the LLM will make the right updates.

Similarly, if a new client relationship needs to be set up — name, invoicing email, account owner, tags — that can also be done through MCP rather than the UI.

Reorganizing work mid-project

Sometimes a project changes shape and you need to move people around. The MCP supports this: move a position from one project to another, move allocations from one position to another, or archive positions that are no longer needed.

This is particularly useful when a deal is won or restructured mid-flight and the staffing plan needs to shift. Instead of manually recreating everything in the UI, you can describe the reorganization and let the LLM handle the steps — while it checks with you before anything that has unexpected implications.

Logging expenses

You can log billable and non-billable expenses directly through MCP, including line items, expense categories, vendors, and the amount to invoice to the client.

For teams that accumulate expenses across multiple projects, you can log several at once by pasting a list and asking the LLM to create the records.

Bulk data entry and corrections

Beyond individual records, the MCP supports writes in bulk. Rather than entering records one at a time in the UI, you can describe what you want to create in conversational or semi-structured form.

This is useful for:

• Pasting a table of time entries and asking the LLM to log them all

• Describing a staffing plan ("allocate Maria 80% to project X and 20% to internal R&D from July to September") and having it create the allocations

• Setting up a complete financial plan for a new project: budgets, planned expenses by category, and rate card assignments in one conversation

The LLM interprets your input and maps it to the Operating data model. That flexibility handles unstructured inputs well, but you should review what was created — especially for anything financial.

What Operating MCP is not

A few things worth setting expectations on.

MCP is a complement to the Operating UI, not a replacement. For day-to-day navigation, visual planning on the timeline, or anything that benefits from seeing data laid out spatially, the UI is faster and clearer. MCP is most useful when you're doing something the UI doesn't make easy, or when you're already in a chat interface and want to stay there.

Write operations are live. Anything the LLM creates or modifies through MCP is real data in your Operating instance — there's no sandbox or draft mode. Notice how in Reorganizing work mid-project section, the model paused to explain the implications and offer options before moving Motoko's position. That's the right behaviour to expect, but you should still review what was created, especially for bulk operations.

It works best with Operating's vocabulary. The closer your descriptions match the concepts in Operating — projects, positions, allocations, people — the more reliably the LLM maps your intent to the right actions. Vague requests produce vague results.