| What | Connect AI assistants, such as
Claude or ChatGPT, to Teamwork
Desk through MCP so support teams
can manage support work using
natural language. |
|---|---|
| Why |
|
| Who | Support teams using Teamwork Desk
with AI assistants such as Claude
or ChatGPT. |
| When | Use the Teamwork.com MCP server
when you want AI assistants to help
review tickets, draft replies,
monitor SLAs, identify support
trends, or manage supported Desk
workflows. |
Use AI assistants with
Teamwork Desk
The Teamwork.com MCP server powers the
connection between AI assistants and Teamwork
Desk. It helps support teams query Desk data,
manage supported workflows, and act on
ticket, customer, SLA, and help doc
information using natural language.
AI assistants can help teams work with
support tickets, customer conversations, help
docs, SLAs, and operational support workflows
using natural language.
By connecting your AI assistant to Teamwork
Desk, teams can spend less time manually
filtering queues, exporting reports, and
chasing updates, and more time responding to
customers and resolving high-priority
work.
Best for support teams managing high
ticket volume, SLA-sensitive
workflows, recurring customer issues,
or large help doc libraries.
What is MCP?
MCP stands for Model Context Protocol. It is
a standard that allows AI assistants, such as
Claude and ChatGPT, to connect to external
tools and services, including Teamwork
Desk.
The Teamwork.com MCP server connects your AI
assistant to Teamwork Desk so it can read,
create, update, and manage supported Desk
workflows using natural language.
On this page
Get help with
everyday Teamwork Desk workflows
The Teamwork.com MCP server helps your AI
assistant work with Teamwork Desk tickets,
customers, SLAs, help documentation, and
support operations workflows using natural
language.
Common support
workflows
- Identify tickets at risk of SLA breach or
waiting too long for a response.
- Triage unassigned or high-priority support
queues.
- Draft contextual replies using ticket
details and help documentation.
- Surface recurring customer issues and help
documentation gaps.
- Review agent activity, first-contact
resolution signals, and operational
trends.
- Create Teamwork.com follow-up work from
support conversations where needed.
Ask for the support outcome you need.
The Teamwork.com MCP server helps
your AI assistant take supported
actions in Desk using the access
available to the connected
account.
Connect AI assistants
to Teamwork Desk
Most MCP-compatible AI assistants follow a
similar setup flow: add the Teamwork.com MCP
server, enter the dedicated server URL,
authenticate with Teamwork.com, and approve
access permissions.
- Enable MCP server access:
- Click your profile icon in your Teamwork Desk account, then select MCP server.
- Toggle on Teamwork Desk MCP Server.
- Add the Teamwork.com MCP
server:
- Open your chosen AI assistant.
- Navigate to the connectors setup area. Depending on the platform, it might be under Apps, Integrations, MCP, or Connectors.
- Create a new app. You might need to request developer mode access from your AI platform admin.
- Enter the Teamwork.com MCP server
URL: https://mcp.ai.teamwork.com.
- Sign in using
the Teamwork.com account you want the AI
assistant to use.
- Review and
approve the requested permissions.
Platform-specific
setup guides
Claude and ChatGPT both use the same
Teamwork.com MCP server URL, though the setup
flow differs by platform. Use the guides below
for platform-specific setup steps. Additional
MCP-compatible AI assistant setup guides may
be added over time as platform support
expands.
Example prompts
Once connected, you can ask Claude or ChatGPT
to help with Teamwork Desk workflows using
prompts like these. Adapt them to the inboxes,
agents, customers, SLAs, and date ranges you
want to review.
Example support
workflow
Find
urgent tickets
“Which
tickets are due to breach SLA in the
next two hours and are
unassigned?”
→
Prioritize
customers
“Which of
those tickets are from high-priority
customers?”
→
Take
action
“Draft a
response acknowledging the delay and
assign the tickets to available
agents.”
More prompts to
try
- Show tickets due to breach SLA in the next
two hours that are currently
unassigned.
- Which tickets in waiting on customer
status have not received a reply in the
last 10 days?
- Which high-priority tickets arrived
outside business hours and waited more
than eight hours for a response?
- Draft replies for new password reset
tickets using our help docs.
- Suggest help doc article topics based on
repeated tickets from the last 30
days.
- Rank agents by ticket replies in the last
30 days and present the results in a
table.
- Find tickets where customers or agents
asked for clarification multiple times in
the last 21 days.
- Find tickets where a task should have been
created but no linked task exists.
Use AI output as a draft or starting point.
Review results before sending customer
replies, assigning tickets, making staffing
decisions, or acting on SLA, ticket, or
customer data.
Supported
Teamwork Desk workflows and data
The Teamwork.com MCP server supports
workflows across Desk data that can be read,
created, updated, or managed through supported
MCP tools and account permissions.
- Business hours
- Companies and customers
- Help doc articles and help doc sites
- SLAs
- Tags, ticket priorities, ticket statuses,
and ticket types
- Tickets and ticket conversations
- User accounts
Available results depend on the information
tracked in Teamwork Desk and the permissions
available to the connected account.
Permissions and
access
The Teamwork.com MCP server uses the
Teamwork.com account connected to the AI
assistant. Available information and actions
depend on the permissions associated with that
account.
Your AI assistant does not get separate
access to Teamwork Desk. It works through the
connected account and follows the permissions
available to that account.
Frequently asked
questions
What is
the Teamwork.com MCP server for
Teamwork Desk?
The Teamwork.com MCP server connects AI
assistants, such as Claude and ChatGPT, to
Teamwork Desk using the Model Context
Protocol (MCP). Once connected, AI
assistants can help search, summarize,
create, and update supported Desk work
using natural language.
Can I use
Claude or ChatGPT with Teamwork
Desk?
Yes. Claude and ChatGPT are supported
through the Teamwork.com MCP server. Use
the setup guides linked above for
platform-specific setup steps.
What can AI
assistants access in Teamwork Desk?
Supported workflows include tickets,
customers, companies, help docs, SLAs,
ticket statuses, priorities, tags,
business hours, users, and other supported
Teamwork Desk data.
Can AI
assistants create and update
tickets?
Yes. Supported actions include creating
and updating tickets and other supported
Teamwork Desk data, based on account
permissions.
Can AI
assistants draft ticket replies?
Yes. AI assistants can help draft
contextual replies using ticket
information and help docs. AI-generated
responses should always be reviewed before
sending.
Does the MCP
server use my Teamwork Desk
permissions?
Yes. AI assistants only have access to
the information and actions available to
the connected Teamwork.com account.