The Jira MCP for teams that run delivery in Jira
Connect one Atlassian account and the issue lifecycle, JQL, worklogs, projects, boards, and sprints become steps an agent or a workflow can run. 93 typed nodes over the Jira Cloud REST API v3 and the Jira Software Agile API.
What the Jira MCP connector does
The Jira connector for Mattering Automations ships 93 typed nodes in five groups. The issue lifecycle: create one issue or up to 50 at once, read one or a hundred, update, assign, transition through the workflow, and read the field-level change history. The conversation around an issue: comments, worklogs, watchers, issue links, links out to things that live outside Jira, and Jira's own email notifications. Search: a bounded JQL query, the same query in a JSON body when it is too long for a URL, an approximate count, and a validator that checks JQL before you run it. The structures work lives in: projects, versions, components, saved filters, users, fields, statuses, priorities, resolutions, and labels. And Jira Software delivery: boards and their column configuration, sprints, epics, the backlog, and ranking.
Each capability is a node on the Automations canvas, and every node speaks MCP, the open protocol for connecting AI assistants to tools. Claude, ChatGPT, and any MCP-capable assistant can use them, and they also power Matter, the platform's built-in AI assistant. Describe the Jira workflow you want in plain language and it gets built on the canvas in front of you, where you can see and edit every step.
Jira ids are not portable: transition ids, issue type ids, and custom field ids differ per project, per workflow, and per site. So the connector ships the discovery reads that make an agent workable without guessing. Jira: List Available Transitions returns the moves an issue can make right now. Jira: List Creatable Issue Types and Jira: List Fields for New Issue return what a specific project will accept. Jira: List Fields maps a custom field name to its customfield_ id, and Jira: Get JQL Reference Data returns the exact JQL vocabulary of your site. Read first, then write.
Two things to know before you build. Every node takes your site URL, for example https://acme.atlassian.net, because Jira Cloud serves each customer from their own host rather than a shared one. And the 18 board, sprint, epic, and backlog nodes call the Jira Software API, so they need a Jira Software licence on the site.
One connection, three surfaces
- Visual automations. Drag Jira nodes onto the canvas and wire them to triggers, conditions, and other apps.
- AI agents, yours or ours. These nodes speak MCP, the open protocol, so Claude, ChatGPT, and any MCP-capable assistant can use them, and Matter, the agent built into the platform, uses them natively.
- The whole platform. Combine with native tasks, projects, and other connectors, no middleware in between.
Every Jira node, ready to drop on the canvas
93 nodes, generated from the connector itself so this list always matches what ships. Expand a category to browse.
Read and query nodes 59
Jira: Count Issues Matching JQL
Get an approximate count of the issues matching a JQL query without fetching them. Much cheaper than a search when you only need the number.
Jira: Find Assignable Users
Find users who can actually be assigned to an issue or in a project. Safer than a plain user search, because not every user is assignable everywhere.
Jira: Get Attachment Details
Get the metadata for one attachment: filename, size, MIME type, author, and the URL its content lives at. The file itself is not returned.
Jira: Get Attachment Settings
Check whether attachments are enabled on the site and what the maximum upload size is, in bytes.
Jira: Get Board
Get one board by id, with its type and the project or user it lives in.
Jira: Get Board Configuration
Get a board's columns, the statuses mapped to each, its filter, and its estimation field. This is how you learn what Done means on that board.
Jira: Get Comment
Get a single comment on an issue by its id, including author, body, and visibility restrictions.
Jira: Get Component
Get one component by id, including its project, lead, and assignee rule.
Jira: Get Current User
Get the Atlassian account the API token belongs to: account id, display name, email, and time zone. Use this to confirm the credential works.
Jira: Get Editable Fields for Issue
List the fields on an issue that the credential can actually edit, with their allowed values. Read this before building an Update Issue body against an unfamiliar project.
Jira: Get Epic
Get one epic by id or key. Gotcha: this endpoint does not work for epics in team managed projects, use Get Issue there instead.
Jira: Get Filter
Get one saved filter by id, including its JQL, owner, and share permissions.
Jira: Get Issue
Get one issue by id or key, for example EX-1. Set fields to a short list to keep the response small, because the default returns every field including custom fields.
Jira: Get Issue History
Get the field change history of an issue, oldest first, with the author and the before and after values of each change.
Jira: Get Issue Link
Get one issue link by id, including both linked issues and the direction of the relationship.
Jira: Get Issues in Bulk
Fetch up to 100 issues by id or key in one request. Issues that could not be read come back in issueErrors rather than failing the call.
Jira: Get JQL Reference Data
Get every field name, operator, and function that JQL accepts on this site, plus the reserved words. This is the vocabulary an agent needs to write valid JQL for a specific site.
Jira: Get Project
Get one project by id or key, including its lead, type, and URL. Project keys are case sensitive.
Jira: Get Site Info
Get information about the Jira site itself: base URL, version, build number, server time, and time zone.
Jira: Get Sprint
Get one sprint by id, with its state, goal, and start and end dates.
Jira: Get User
Get one user by Atlassian account id. Email may be hidden depending on that user's privacy settings.
Jira: Get Version
Get one version by id, including whether it is released, archived, or overdue.
Jira: List All Users
List every user on the site, active and inactive, page by page. Capped at 1000 per page and needs the Browse Users and Groups permission.
Jira: List Available Transitions
List the workflow transitions the issue can take right now, with their ids and target statuses. Transition ids are workflow specific, so read them here before moving an issue.
Jira: List Board Backlog
List the backlog of a board, meaning incomplete issues not in any active or future sprint. Paging uses nextPageToken.
Jira: List Board Epics
List the epics on a board, with their colour and done state.
Jira: List Board Issues
List the issues on a board, including Jira Software fields such as sprint and epic. Paging uses nextPageToken, not startAt.
Jira: List Board Projects
List the projects a board draws issues from. A board can span several projects when its filter allows it.
Jira: List Board Sprints
List the sprints on a board. Filter by state to get just the active one, which is the usual starting point for a standup automation.
Jira: List Boards
List the Jira Software boards the credential can view, optionally filtered by type or project. Requires a Jira Software licence on the site.
Jira: List Creatable Issue Types
List the issue types the credential can create in a project, with their ids. Issue type ids differ per project, so read them here rather than assuming them.
Jira: List Epic Issues
List the issues under an epic, ordered by rank. For team managed projects, search with JQL on parent instead.
Jira: List Fields
List every system and custom field on the site with its id, name, and JQL clause names. This is how you map a custom field name to its customfield_ id.
Jira: List Fields for New Issue
List the fields available when creating one issue type in one project, including which are required and what values they accept. This is how you discover custom field ids.
Jira: List Groups for User
List the groups a user belongs to, by Atlassian account id. Useful for permission checks before an automation acts on someone's behalf.
Jira: List Issue Comments
List the comments on an issue. Comment bodies come back as Atlassian Document Format, so expand renderedBody if you want HTML instead.
Jira: List Issue Link Types
List the link types configured on the site, such as Blocks or Duplicate, with their inward and outward wording. Needed before creating a link.
Jira: List Issue Types
List all issue types visible to the credential, across every project it can browse.
Jira: List Issue Types for Project
List the issue types in one project. Takes the numeric project id, not the project key.
Jira: List Issue Watchers
List who is watching an issue and how many watchers it has.
Jira: List Issue Worklogs
List the time logged against an issue, oldest first. Time tracking has to be enabled on the site or this returns an error.
Jira: List Labels
List every label in use on the site, page by page. Labels are free text in Jira, so this is the only way to see what already exists.
Jira: List My Filters
List the filters owned by the credential's user, and optionally their favourites too.
Jira: List Priorities
List the issue priorities on the site, optionally narrowed to the ones available in given projects.
Jira: List Project Components
List every component in a project, with its lead and default assignee rule. Not paginated.
Jira: List Project Roles
List the project roles and the URL for each role's members. Roles are shared across all projects on a Jira Cloud site.
Jira: List Project Statuses
List the valid statuses in a project, grouped by issue type. Statuses differ per issue type, which is why they are nested.
Jira: List Project Versions
List every version, meaning release, in a project with its release date and released flag. Not paginated.
Jira: List Recent Projects
List up to 20 projects the credential's user viewed most recently. Handy as a cheap orientation call.
Jira: List Remote Links
List the links from an issue to things outside Jira, such as a support ticket or a document.
Jira: List Resolutions
List the resolutions on the site, such as Done or Won't Do. Needed when a transition screen requires a resolution.
Jira: List Sprint Issues
List the issues in a sprint, ordered by rank, including Jira Software fields such as story points and flagged. Paging uses nextPageToken.
Jira: List Statuses
List every status used by an active workflow on the site, with its status category.
Jira: Search Filters
Find saved filters by name, owner, or project. Expand jql if you want the query text, it is not included by default.
Jira: Search Issues with JQL
Search issues with a JQL query. Two gotchas: the query must be bounded, so include a filter such as project or updated, and paging uses nextPageToken from the previous response rather than an offset.
Jira: Search Issues with JQL, JSON Body
Same search as the JQL search node, but the query travels in a JSON body. Use this when the JQL or the field list is too long to fit in a URL.
Jira: Search Projects
List the projects the credential can see, with paging and filters. Use this rather than assuming project keys.
Jira: Search Users
Find active users by name or email fragment. Needs the Browse Users and Groups permission, and returns account ids you can use to assign issues.
Jira: Validate JQL Query
Parse and validate JQL queries before running them, returning the parsed structure or the errors. Useful when an agent composes JQL and you want to fail early instead of on the search.
Action nodes 34
Jira: Add Comment
Add a comment to an issue. Gotcha: v3 requires Atlassian Document Format, a plain string body is rejected.
Jira: Add Issue Watcher
Add a user as a watcher on an issue. Gotcha: the body is a bare JSON string holding the account id, not an object.
Jira: Add Remote Link
Link an issue to an external URL. Gotcha: this is an upsert, a call carrying an existing globalId overwrites that link and blanks any field you leave out.
Jira: Assign Issue
Set the assignee on an issue by Atlassian account id. Works even for users who only hold the Assign Issues permission, not full edit rights.
Jira: Create Component
Create a component in a project. Components group issues inside a project and can set a default assignee.
Jira: Create Filter
Save a JQL query as a reusable filter. It is created with the site's default share scope and is not favourited.
Jira: Create Issue
Create an issue or subtask. Gotcha: description and other rich text fields must be Atlassian Document Format, and the response returns only the new id, key, and self URL, not the whole issue.
Jira: Create Issues in Bulk
Create up to 50 issues in one call. Partial success is normal: check the errors array in the response, because valid issues are still created when others fail.
Jira: Create Project
Create a project from a project template. Needs Jira administrator rights, and the template key has to match the project type key.
Jira: Create Sprint
Create a future sprint on a scrum board. It is created in the future state, use Update Sprint to start it.
Jira: Create Version
Create a version, meaning a release, in a project. Note that projectId is numeric, so look it up with Get Project if you only have the key.
Jira: Delete Attachment
Permanently delete an attachment from an issue. Deleting another user's attachment needs the Delete All Attachments permission.
Jira: Delete Comment
Permanently delete a comment from an issue. Deleting another user's comment needs the Delete All Comments permission.
Jira: Delete Component
Delete a component. Set moveIssuesTo to hand its issues to another component, otherwise they are simply left without one.
Jira: Delete Issue
Permanently delete an issue. An issue that has subtasks cannot be deleted unless deleteSubtasks is true, which deletes those too.
Jira: Delete Issue Link
Remove the link between two issues. The issues themselves are untouched.
Jira: Delete Project
Delete a project and everything in it. Needs Jira administrator rights, and an archived project has to be restored before it can be deleted.
Jira: Delete Worklog
Permanently delete a worklog entry. Use adjustEstimate and increaseBy if the remaining estimate should be given the time back.
Jira: Link Two Issues
Create a link between two issues, for example blocks or duplicates. Gotcha: the response body is empty, so read the issue back if you need the new link id.
Jira: Log Work
Log time against an issue. By default Jira reduces the remaining estimate automatically, use adjustEstimate to change that.
Jira: Move Issues to Backlog
Move up to 50 issues back to the backlog, which removes them from any future or active sprint.
Jira: Move Issues to Epic
Move up to 50 issues under an epic. An issue can only sit in one epic, so this removes it from any previous one.
Jira: Move Issues to Sprint
Move up to 50 issues into a sprint. Only future and active sprints accept issues, a closed sprint is rejected.
Jira: Rank Issues
Reorder issues by ranking them before or after another issue. Gotcha: a partial failure returns HTTP 207 with a per issue status list rather than a plain error.
Jira: Remove Issue Watcher
Stop a user watching an issue. Removing anyone other than yourself needs the Manage Watchers permission.
Jira: Send Issue Notification
Queue a Jira email notification about an issue to its reporter, assignee, watchers, or named users. The mail is sent by Jira, so it comes from the site's own notification address.
Jira: Transition Issue
Move an issue to a new status by transition id. Get the id from List Available Transitions first, because the same status has different ids in different workflows.
Jira: Update Comment
Replace the text of an existing comment. The whole body is overwritten, so send the full new text rather than a fragment.
Jira: Update Component
Update a component's name, description, lead, or assignee rule. Fields you send are overwritten outright.
Jira: Update Issue
Edit fields on an issue. Gotcha: this cannot change status, use Transition Issue for that, and the response is empty unless you set returnIssue.
Jira: Update Project
Update project details such as name, description, lead, or URL. Changing the key or a scheme needs Jira administrator rights.
Jira: Update Sprint
Partially update a sprint, including starting or completing it through the state field. On a closed sprint only the name and goal can still change.
Jira: Update Version
Update a version, for example to mark it released or move its release date. This is how an automation closes out a release.
Jira: Update Worklog
Change an existing worklog entry, such as its duration, start time, or comment.
Connecting Jira takes one credential
Jira Cloud authenticates with two values: the email address of an Atlassian account and an API token created for it, combined into a standard HTTP Basic header. Atlassian deprecated password authentication, so a token is the credential that works. Create one at id.atlassian.com under Account settings, Security, API tokens. The token carries exactly the Jira permissions of the account it belongs to, with no separate scoping, so issue it from an account whose access matches what your automations should be able to do.
In Atlassian, open Account settings, Security, API tokens, and create a token. Copy it before you close the dialog, because Atlassian shows it once.
In Mattering Automations, open Connectors, choose Jira, and create a credential: your Atlassian account email goes in the username field, the API token goes in the password field.
Fill in the site URL on each Jira node, for example https://acme.atlassian.net. It is required on every node, because Jira Cloud has no shared API host.
There is no Test Connection button for this connector, since a fixed-URL probe has nothing to probe against. Verify by dropping a Jira: Get Current User node on a canvas and running it with just the site URL. A successful response names the account the token belongs to.
Common uses for the Jira MCP
Triage the queue instead of reading it
Run Jira: Search Issues with JQL on a schedule for unassigned bugs, stale tickets, or anything raised in the last day, then let an agent read each one, add a comment, set the assignee, and move it through the workflow with Jira: Transition Issue.
Standups and sprint reporting that write themselves
Jira: List Board Sprints filtered to the active sprint, then Jira: List Sprint Issues, gives you the whole board state in two steps. Summarize it with an AI step and post it to a message, an email, or a Mattering Productivity task before the meeting starts.
Turn intake into issues
A form submission, a support email, or a row in a Database becomes a Jira issue: Jira: Create Issue for one, Jira: Create Issues in Bulk for a batch, Jira: Link Two Issues when it duplicates something already open, and Jira: Add Issue Watcher so the requester follows it.
Close out a release without the checklist
Jira: List Project Versions and Jira: Search Issues with JQL tell you what is still open against a fix version. When it is clear, Jira: Update Version marks the release done and the same flow can announce it wherever your team reads announcements.
Time tracking rollups
Jira: List Issue Worklogs across a project or a sprint gives you logged time by person and by issue, ready for an invoice draft or a weekly report. Jira: Log Work writes time back when an automation completes work of its own.
Backlog grooming an agent can actually do
Jira: Move Issues to Sprint, Jira: Move Issues to Epic, Jira: Move Issues to Backlog, and Jira: Rank Issues handle up to 50 issues per call, so an agent can reshape a sprint from a plain-language instruction rather than a morning of dragging cards.
Example: A new support request becomes a triaged Jira issue
Built on the Automations canvas in a few minutes, or described in plain language to Matter and built for you.
Trigger: a scheduled step or an inbound webhook brings a new support request onto the canvas.
Jira: Search Issues with JQL checks the project for an open issue that already covers it.
Jira: Create Issue opens one when nothing matches, with the summary and an Atlassian Document Format description.
Jira: Assign Issue sets the owner and Jira: Add Issue Watcher keeps the requester in the loop.
Jira: List Available Transitions reads the workflow, then Jira: Transition Issue moves the issue into triage.
Mattering Productivity: a task is created carrying the new issue key, so the work is tracked on both sides.
Jira MCP questions, answered
Can an AI agent create, update, and transition Jira issues through this MCP?
Yes. Matter and any MCP-connected assistant, including Claude and ChatGPT, can create issues one at a time or in bulk, edit fields, assign, comment, log work, link issues, add watchers, and move issues through the workflow. Status is the one thing Update Issue cannot change: Jira requires a transition for that, so read Jira: List Available Transitions and then run Jira: Transition Issue.
Does this work with Claude, ChatGPT, or other AI assistants?
Yes. Mattering exposes these nodes over MCP, an open protocol for connecting AI assistants to tools, so any MCP-capable assistant, including Claude and ChatGPT, can connect to your workspace and work with Jira through them. Matter, the assistant built into the platform, uses the same nodes natively.
What credential does the connector use?
An Atlassian account email address plus an API token, sent together as a standard HTTP Basic header. It is not a bearer token, and an Atlassian password will not work, because Atlassian deprecated password authentication for the REST API. You create the token at id.atlassian.com and Mattering stores it encrypted in Google Secret Manager. The token inherits the Jira permissions of the account that owns it, so the connector can reach exactly what that person can reach.
Why does every Jira node ask for a site URL?
Because Jira Cloud has no shared API host. Every customer is served from their own site, such as https://acme.atlassian.net, or from a custom domain, so there is no single address the connector could assume. Putting the host on the node rather than on the credential also means one credential can drive two sites, for example a production site and a sandbox, from the same canvas. The field accepts a bare host or a full URL and normalizes it.
Do the board, sprint, and epic nodes work on any Jira site?
The 18 board, sprint, epic, and backlog nodes call the Jira Software API, so the site needs a Jira Software licence. Boards and sprints do not exist on a site that only has Jira Work Management. Two of those nodes carry a further limit that is worth knowing up front: Atlassian documents Jira: Get Epic and Jira: List Epic Issues as not working for team managed (next generation) projects, which is what most new Jira Cloud projects are. In a team managed project, read the epic with Jira: Get Issue and list its children with a JQL search on parent instead.
Why do comments and descriptions need Atlassian Document Format?
REST API v3 takes rich text as an Atlassian Document Format object, not a plain string and not wiki markup, so a description, a comment body, or a worklog comment sent as text is rejected. The affected nodes say so in their field descriptions and show the shape. It is the most common Jira API mistake and it is worth handling once in your first flow.
Can a Jira event start a Mattering workflow?
Not directly from this connector: it has no trigger nodes, and Atlassian restricts its webhook API to Connect and OAuth 2.0 apps, so an API token cannot register one. Two ways around it. Create a Jira automation rule inside Jira that posts to a Mattering webhook trigger, or poll on a schedule with Jira: Search Issues with JQL filtered on updated and process what changed since the last run.
What do the Jira nodes not cover?
Attachment upload is not included, because Jira takes it as a multipart file post and a node has no file to post. Jira: Get Attachment Details returns an attachment's metadata and the URL its content lives at, which an HTTP step can fetch. Site administration schemes (field configurations, screens, permission and workflow schemes) are deliberately out of scope: they are a way to break a customer's Jira from an automation, not a work management surface. Creating and deleting Atlassian accounts is out of scope for the same reason.
Put Jira to work inside your workspace
Connect once, then use every node in visual automations and AI agent workflows across the Mattering platform.
All product names, logos, and brands are property of their respective owners and are used for identification purposes only. Use of these names, logos, and brands does not imply endorsement or affiliation. Jira is a product of its respective owner; Mattering is an independent platform.