Skip to content
Tallyfy
Developers

Using Tallyfy MCP server with Microsoft Copilot Studio

Run your Tallyfy work from Microsoft Copilot

Section titled “Run your Tallyfy work from Microsoft Copilot”

You can run Tallyfy by talking to Microsoft Copilot. Ask it to build a template from your own notes, update tasks after a meeting, or tell you which processes are running late, and it does the work in Tallyfy for you.

You set this up once in Copilot Studio, Microsoft’s tool for building your own Copilot agent. After that, you and the people you share it with use the agent in Copilot chat and in Teams.

The ordinary Copilot chat can’t add Tallyfy on its own, because it has no place to type in an MCP1 server address. That’s why setup goes through Copilot Studio.

What you can do

Section titled “What you can do”

Some things people ask for:

  • Paste a checklist from a Word document and ask for a new template with one step per item.
  • Paste your notes from a planning meeting and ask Copilot to add the steps the template is missing.
  • Paste a contractor’s meeting notes and ask which tasks in the build process they’ve finished.
  • Ask which processes are running late this week, and who’s holding them up.
  • Ask to move a task from one person to another while someone is away.

What you need

Section titled “What you need”
  • Access to Copilot Studio. Your IT team decides who can build agents there.
  • A license that can publish. The person who builds the agent needs a Microsoft 365 Copilot license, or a Copilot Studio user license with credits for the environment. Trial licenses can’t publish an agent like this one, according to Microsoft’s licensing FAQ ↗[1].
  • A Tallyfy account in a workspace where AI connections are turned on. If they aren’t, Tallyfy’s sign-in page tells you so and gives you a ready-made email to ask Tallyfy support to add them.

Set it up

Section titled “Set it up”

  1. Create an agent. Open Copilot Studio ↗[2]. On the home page, select Other ways to build, then create a new agent. If your screen shows a New experience switch instead, turn it off, open Agents, and select Create blank agent. A name like “Tallyfy assistant” works well.

    Other ways to build gives you a standard agent, which is what these steps follow. The cards at the top of the home page build Microsoft’s newer GitHub Copilot-based agents instead, which have different menus and different billing (see Microsoft’s Start building guide ↗[3] and the Cost section below).

  2. Add Tallyfy as a tool. On the agent’s Tools page, select Add a tool, then New tool, then Model Context Protocol.

  3. Fill in the server details.

    • Server name: Tallyfy
    • Server URL: https://mcp.tallyfy.com
    • Server description: paste the text below. Copilot reads it to decide when to use Tallyfy, so keep it specific.
    Tallyfy runs our business processes. Use it to find, create and edit templates, steps, form fields and automation rules, to launch and track processes, and to read, complete or reassign tasks.

  4. Choose how to sign in. Select OAuth 2.02, then Dynamic discovery, then Create, then Next. Copilot finds Tallyfy’s sign-in details by itself, so there’s no key or secret to paste.

  5. Connect your Tallyfy account. In the Add tool box, select Create a new connection. A Tallyfy window opens. Sign in, pick your organization if you belong to more than one, and tick the permissions Copilot should have.

    To let Copilot build and edit templates, tick Create and manage your templates. Tick Create and manage form fields and Create and manage automation rules too if you want it to add fields and rules. Then select Authorize, and back in Copilot Studio select Add to agent.

  6. Keep each person’s own sign-in. In the Tallyfy tool’s settings, leave Authentication on End user. That’s Microsoft’s default, and it means everyone who uses the agent signs in to Tallyfy as themselves. Maker-provided would make every user act as you.

  7. Pick the tools you need. Tallyfy offers more than 100 tools, and Microsoft’s tools guide ↗[4] suggests keeping an agent to 25 to 30 tools for the best results. Every tool starts switched on. Turn off Allow all, then turn off each tool you don’t need, so you’re left with the ones for the job. For building templates, keep the tools in the Templates and steps section of the tool list, which is about two dozen, and add a few form field tools if you want Copilot to add fields. Once Allow all is off, tools Tallyfy adds later start switched off.

  8. Test it. In the test chat, ask List my templates in Tallyfy. You should see your Tallyfy templates.

  9. Put it in Copilot and Teams. Select Publish. Then go to Channels, open Teams and Microsoft Copilot, make sure Make agent available in Microsoft Copilot is selected, and select Add channel. Then select See agent in Teams and Add to install it for yourself. This needs your organization to allow Power Platform apps in Teams; if Add fails, ask your Teams admin.

  10. Share it with colleagues. Select the three dots (…) at the top of the agent, select Share, add a security group, and choose User - can use the agent so they can chat with it without editing it. Microsoft’s sharing guide lists what they need, as noted under Sharing in the IT section below. Then go back to Channels > Teams and Microsoft Copilot, select Availability options, select Show to my teammates and shared users, check that Show in Built By Your Colleagues is selected, and select Update. The people you shared it with will find it in the Built with Power Platform section of the Teams app store, or you can send them the install link from Availability options > Copy link (the link doesn’t work in the Teams mobile app). Showing it to everyone in the company is a separate process with admin approval, covered in the publishing guide. Microsoft’s publishing guide ↗[5] has the full details.

If something goes wrong

Section titled “If something goes wrong”
  • A Tallyfy page says AI connectors aren’t on your plan. AI connections aren’t turned on for your Tallyfy workspace yet. Select Email support to add it on that page and Tallyfy support takes it from there.
  • Tallyfy’s permission screen says “Unverified third-party application”. You’ll see this when you connect Copilot today, because Copilot Studio’s sign-in address isn’t on Tallyfy’s list of known apps yet. Continue only if you started the connection yourself a moment ago, from Copilot Studio, Teams or Copilot.
  • Copilot can read but won’t change anything. There are three usual causes. The Tallyfy sign-in it’s using may lack the write permission, such as Create and manage your templates: remove that Tallyfy connection, sign in again and tick it. The tool it needs may be switched off in the agent’s tool list. Or your own Tallyfy role may not allow the change, since Copilot can’t do more than you can. If you didn’t build the agent, ask the person who did.
  • Publishing fails with a license message. The builder is probably on a trial license. Ask your IT team for one of the licenses under What you need.
  • The agent never calls Tallyfy. MCP tools need generative orchestration, which is on by default for new agents. If you reused an older agent, check it’s turned on in the agent’s settings.
  • Signing in or loading the tools fails inside Copilot Studio. Check Microsoft’s MCP troubleshooting page ↗[6], then contact Tallyfy support with the time it failed and the error you saw.

Is it secure?

Section titled “Is it secure?”

Yes. Two separate sign-ins are involved. You get into Copilot Studio with your normal Microsoft account. Copilot then connects to Tallyfy with a separate Tallyfy sign-in. Copilot doesn’t pass your Microsoft sign-in to Tallyfy, so you sign in to Tallyfy the way you normally do.

Each person who uses the agent signs in with their own Tallyfy account, so Copilot works inside that person’s Tallyfy permissions.

For your IT team

Section titled “For your IT team”

(Skip this unless you’re setting up the technical side.)

Keeping it to the people who need it

Section titled “Keeping it to the people who need it”

The agent reaches only the people it’s shared with, so you can start with one person and widen it later. A few settings decide who gets what:

  • Data policies. Copilot Studio reaches the Tallyfy server through a Power Platform connector, so your data policies apply to it, and blocking that connector blocks the Tallyfy tools (Microsoft’s data policy guide ↗[7]). If your policy’s Custom connectors page tags the * pattern as Blocked, add a pattern for https://mcp.tallyfy.com* above it and tag it Business (custom connector patterns ↗[8]). Business works only if the agent’s other connectors are in the Business group too, because connectors in different groups can’t share data.
  • Sharing. The agent’s Share option gives named people or security groups access (Microsoft’s sharing guide ↗[9]). Listing it for the whole company in the Teams app store needs your admin’s approval. The same guide says the people you share with need a Copilot Studio per user license and the Environment Maker role in the environment, which an environment administrator assigns. Check those before you widen access.

Cost

Section titled “Cost”

Microsoft’s billing rates table ↗[10] lists generative answers and agent actions, the features this setup uses, as “No charge” when a person with a Microsoft 365 Copilot license uses the agent under their own identity in Microsoft 365, such as in Teams or Copilot. People without that license use Copilot Credits, and so do agent flows started by other triggers. Check the conditions on that page for your tenant.

That applies to standard agents, which step 1 creates. Microsoft’s newer GitHub Copilot-based agents use consumption-based billing: building, testing and running them can use Copilot Credits that a Microsoft 365 Copilot license doesn’t cover, according to Microsoft’s cost guide for those agents ↗[11].

To set a hard ceiling, open the Power Platform admin center, go to Licensing > Copilot Studio > Manage Agents, select the agent, select Set limit, enter a monthly credit limit, and turn on Stop usage. Without Stop usage, the limit only sends alerts. The list shows only agents that have incurred charges, so a new agent may not appear there yet. To cap a new agent before it appears there, give its environment a fixed amount of Copilot Credits under Manage Copilot Credits and turn off Draw from the available capacity in my tenant. If the environment is already linked to a pay-as-you-go billing plan, unlink it unless you’ve decided to use it. Microsoft’s credit limits guide ↗[12] and its cost guide for GitHub Copilot-based agents (linked above) walk through these settings.

Copilot Cowork is a different product and is billed per use. This setup doesn’t need it.

On the Tallyfy side, AI connections have to be on your Tallyfy plan. If they aren’t, the sign-in page offers Tallyfy’s AI connector add-on and shows its price.

Data and compliance

Section titled “Data and compliance”

Copilot sends Tallyfy what each person asks about, and Tallyfy sends back what that person’s account can see. Once data reaches Tallyfy, your agreement with Tallyfy covers it. Microsoft’s agent settings guide ↗[13] says data processed by non-Microsoft services isn’t subject to Microsoft agreements. Tallyfy’s security and compliance details are on the compliance page.

Tallyfy doesn’t block anyone from typing sensitive data into a form field or a comment. If you handle regulated data, decide what belongs in Tallyfy before you widen access.

For developers

Section titled “For developers”

(Skip this unless you’re setting up the technical side.)

Server endpoint and support status

Section titled “Server endpoint and support status”

The MCP server is hosted at https://mcp.tallyfy.com and exposes the full Tallyfy tool set over streamable HTTP. https://mcp.tallyfy.com/mcp reaches the same server. It’s listed on the Official MCP Registry ↗[14] as com.tallyfy/mcp-server. The full list of tools is on Every tool the MCP server gives your AI.

Tallyfy also has a Power Automate connector ↗[15] listed by Microsoft. The MCP server is added through Copilot Studio’s own MCP wizard, so you don’t need to build a custom connector or write an OpenAPI file.

How sign-in works

Section titled “How sign-in works”

Copilot Studio’s Dynamic discovery option reads Tallyfy’s discovery documents, registers itself as a client, and runs the authorization code flow. You don’t configure any of it by hand.

ItemValue
Protected resource metadatahttps://mcp.tallyfy.com/.well-known/oauth-protected-resource
Authorization server metadatahttps://mcp.tallyfy.com/.well-known/oauth-authorization-server
Client registrationhttps://mcp.tallyfy.com/mcp/oauth/register
Authorizationhttps://mcp.tallyfy.com/mcp/oauth/authorize
Tokenhttps://mcp.tallyfy.com/mcp/oauth/token
Grant typesauthorization_code, refresh_token
PKCES256, required
Client authenticationnone, client_secret_post, client_secret_basic
Access token lifetime1 hour
Refresh token lifetime30 days

The twelve scopes follow mcp.{resource}.{action}, with read and write for each of users, tasks, processes, templates, forms and automation. The person signing in chooses which ones to grant on Tallyfy’s consent screen, whatever the client asked for. A normal MCP request with no valid bearer token gets 401 with a WWW-Authenticate header that points at the protected resource metadata.

Three tools, ask_user_question, ask_user_to_rank and ask_user_to_confirm, are built for Tallyfy’s own chat window, which shows their answer as a form. Turn them off in Copilot.

Multi-agent orchestration

Section titled “Multi-agent orchestration”

Copilot Studio can coordinate multiple specialized agents for workflows like employee onboarding.

Multi-agent orchestrationUser Request(New Employee)Primary Agent(Orchestrator)AnalyzeRequirementsHR AgentIT AgentTraining AgentProgressMonitorTallyfyTrackerOnboardingComplete InitiateonboardingParse requestCreate HRtasksCreate ITtasksScheduletrainingReportprogressReportprogressReportprogressUpdatestatusProcesscomplete

What to notice:

  • The Primary Agent acts as an orchestrator, delegating to specialized agents
  • Each agent creates and manages its tasks through Tallyfy’s API
  • All progress is monitored through Tallyfy’s tracker view

A few worked examples of agents in action:

  • Multi-agent onboarding. Prompt: “Start the employee onboarding process for Jane Smith, assign tasks to HR and IT, and schedule training sessions.” The primary agent receives the request, delegates documentation tasks to the HR agent (via Tallyfy), system-access setup to the IT agent, and session scheduling to the Training agent, monitors progress across all of them, and reports a consolidated status back to you.
  • Power Platform integration. Prompt: “When a new sales opportunity is created in Dynamics 365, start our proposal process in Tallyfy and create a Teams channel for collaboration.” A Power Automate flow triggers on the Dynamics 365 opportunity, the MCP server creates a Tallyfy process instance, the Teams connector creates a dedicated channel, and the agent monitors and reports progress.
  • Enterprise compliance. Prompt: “Review all completed audit processes this quarter and generate a compliance report with any violations highlighted.” The agent queries Tallyfy for audit process data, analyzes completion times against SLAs, checks for required approvals, generates a formatted report in SharePoint, and routes it to the compliance team if violations are found.

More example prompts that work well here:

  • “When a purchase order exceeds $50,000 in Dynamics 365, initiate our approval workflow in Tallyfy, notify team leads in Teams, and track in Power BI.”
  • “Monitor all SOX compliance processes, ensure proper segregation of duties, and generate audit reports with full trail documentation.”
  • “Coordinate our product launch process across marketing, sales, and operations teams, with each department’s agent handling their specific tasks.”
  • “Analyze our customer service processes to identify bottlenecks, suggest optimizations, and automatically implement approved changes.”
  • “Manage our FDA submission process with full audit trails, approval chains, and automatic compliance checking at each step.”

Multi-agent setups delegate tasks between specialized agents, keep context across agent handoffs, give you central monitoring of the whole workflow, and include built-in error handling and fallback.

Power Platform features

Section titled “Power Platform features”
  • Power Automate flows - trigger workflows from MCP actions
  • Power Apps integration - build custom UIs for Tallyfy data
  • Dataverse storage - store workflow metadata in Microsoft’s database
  • Power BI reporting - create dashboards from Tallyfy analytics

Copilot Studio also gives you a tools and prompts framework: a prompt library to create and share custom prompts across agents, a tool catalog to browse and install prebuilt MCP connectors, version control to track changes to tools and prompts, and a testing framework to validate tools before deployment.

Section titled “Related articles”

Footnotes

Section titled “Footnotes”

  1. Model Context Protocol, the open standard AI assistants use to connect to apps like Tallyfy. ↩

  2. OAuth is the standard that lets you allow one app to act in another for you, without giving it your password. ↩

Get in touch
About Tallyfy