Hi community,
Your goals and projects hold some of the most valuable context in your organization - what you're aiming for, who owns it, how it's tracking, and how it connects to everything else. Goals and Projects tools are now available on the Atlassian MCP Server, so any MCP-compatible client or agent - Claude, ChatGPT, Gemini, or something you've built in-house - can search, read, create, and post updates on goals and projects on your behalf.
Remind me, what is MCP?
The Model Context Protocol (MCP) is an open standard that lets AI agents connect to the systems where your work actually lives. The Atlassian MCP Server exposes Atlassian apps as a set of tools an agent can call - with your permissions, so an agent only ever sees and changes what you can.
What's available
There are four toolsets - a read and a write set for each app. The supported tools documentation provides the full list of tooling, including parameters, scopes, and setup details. Here’s an overview of what’s available:
Goals tools
read_goals
searchGoals - find goals by text and structured filtersgetGoal - full detail on one specific goal, plus related records if you want themgetGoalUpdate - full detail on one specific goal updategetGoalTypes - the goal types a site has enabled for creating goals and success measures
write_goals
createGoal - create a goal or success measurecreateGoalUpdate - post a progress update on an existing goaladdConnectionsToGoals - add a connection to one or more goals
Projects tools
read_projects
searchProjects - find projects by text and structured filtersgetProject - full detail on one specific project, plus related records if you want themgetProjectUpdate - full detail on one specific project update
write_projects
- createProject - create an Atlassian project
- createProjectUpdate - post a progress update on an existing project
- addConnectionsToProjects - add a connection to one or more projects
For the full list of Atlassian’s MCP tooling, including Jira and Confluence, check out our Supported tools documentation.
What can you actually do with it?
Here are some example workflows for you to try - all of these are things you can ask an agent in plain language.
Turn a planning conversation into real, tracked work
"Take the notes from my planning doc and create a goal for improving onboarding activation, owned by me, targeting June 2027 - plus a success measure for activation rate."
The agent checks getGoalTypes, creates the goal and its success measure, and comes back with a link.
Stop creating duplicates
"Before I create anything - what goals and projects already exist around customer onboarding?"
searchGoals and searchProjects let an agent look first and create second.
Draft status updates from the evidence, not from memory
"Here are my notes from this week's sync. Find the matching projects and post an update on each one."
The agent searches for the right records, distils the notes into a clean update, and posts via createProjectUpdate or createGoalUpdate. Pair it with the Jira tools and it can pull the delivery evidence itself before writing anything.
Answer portfolio questions
"Which goals in my org are off track, who owns them, and what was the last update on each?"
searchGoals with structured filters plus getGoalUpdate gives an agent enough to assemble a real readout - no dashboard screenshotting required.
Keep your structure tidy as work grows
"Tag every project connected to the FY27 platform goal with 'platform-modernisation' and add the platform team as a follower."
One addConnectionsToProjects call per connection type, across many projects at once.
Spin up a full initiative in one pass
"Create a project for the billing migration, create a goal for reducing billing support tickets by 30%, connect them, and put both under our Payments focus area."
This is where it gets fun - the agent chains create and connect calls so the initiative arrives fully wired up, instead of as an empty shell someone has to go and link later.
Build your own agents on top
Because these are standard MCP tools, they're not just for chat. You can wire them into scheduled agents that post weekly rollups, into intake flows that create a goal whenever a proposal is approved, or into your own internal assistants.
How to get started
Full instructions are in Get started with the Atlassian MCP Server. The short version:
- Connect your client to https://mcp.atlassian.com/v2/mcp. Most coding agents can install and configure themselves - the doc has a prompt you can paste straight in.
- Sign in when your client starts the Atlassian authentication flow. OAuth 2.1 means every call respects the permissions you already have.
- Ask a question to test it out and confirm that it’s wired up correctly.
One thing to be aware of: tool calls through the Atlassian MCP Server consume Rovo credits. See How Rovo credits work for more information.
Continuing our momentum
We are continuously expanding our MCP capabilities to support deeper, connected workflows. Upcoming additions include expanded coverage for Goals and Projects with broader read surfaces, richer filtering, and write/update capabilities, including metrics and metric values, risks, dependencies, and focus areas.
Also see the recent article about Strategy Collection tools now available on the Atlassian Rovo MCP Server.
We'd love to hear from you
We’re excited for you to begin using these new tools, and would also love to learn more about what you want to do with them. So please tell us in the comments:
- What's the first workflow you'd automate with the Goals and Projects tools?
- What's missing from this first set of tools that would unblock a real use case for you?
- Where would you not want an agent acting on your goals and projects? Boundaries are as useful to us as wish lists.
If you would be interested in future research, usability testing, or feedback sessions for Goals and Projects AI workflows, let us know in the comments as well.
We read every reply, and this feedback goes straight into what we prioritise next.
As always, thank you for being a part of the Atlassian community!
