MCP servers that can create an issue

5 verified servers expose a tool that can create an issue in a tracker

An agent that finds a bug mid-task needs to file it where a human will see it, not bury it in a chat log. Every major tracker with an MCP server exposes one tool for that, under a different name each time.

These verified servers let an agent create an issue, with the exact tool each tracker uses.

Top pick

GitHub

GitHub

Official

GitHub's official remote MCP server for repos, issues, pull requests, Actions, and code search.

version-control30,334
Tool:
  • issue_write

GitHub folds create and edit into a single issue_write call, sitting beside the comment and pull-request tools the same agent reaches for next.

Pick 2

Linear

Linear

Official

Linear's official remote MCP server lets agents create, search, and update issues and projects.

productivity
Tools:
  • create_issue
  • create_issue_label

Two separate tools, used in sequence: create_issue_label mints a missing label, then create_issue files against it, so an agent-filed ticket need not land untriaged.

Pick 3

Atlassian (Jira & Confluence)

Atlassian

Official

Atlassian's official remote MCP server for Jira and Confluence: read, search, create, and update issues and pages.

project-management746
Tool:
  • createJiraIssue

Jira demands a project and an issue type on every create, and createJiraIssue supplies both through the official Atlassian server.

Pick 4

GitLab

GitLab

Official

GitLab's official, built-in remote MCP server for issues, merge requests, pipelines, and code search via OAuth.

version-control
Tool:
  • create_issue

A plain create_issue keeps the agent on GitLab, in the same place as the repository and the pipelines it just touched.

Pick 5

Gitea

Gitea

Official

Gitea's official MCP server for repos, branches, issues, pull request reviews, releases, Actions, and wikis on any Gitea instance.

version-control73
Tools:
  • create_issue
  • create_issue_comment

Self-hosted forges are covered too: Gitea creates the issue and can follow up with a comment, all on infrastructure you run.

What to know

The interesting differences are in the fields. A bare title-and-body create is the floor; what varies is whether the agent can set a label, assignee, project, or issue type in the same call, or has to follow up with a second tool. Linear and GitHub lean rich here, which matters if you want the agent to file something triaged rather than a naked title a human still has to sort.

The quiet failure mode is duplicates: an agent with no record of what it already filed opens the same issue twice across two sessions. A tidy backlog depends on the agent knowing which tickets it has already created, and nothing in the create call itself tells it.

Questions

Why is GitHub's tool called issue_write and not create_issue?
The GitHub server folds creating and editing an issue into a single issue_write tool rather than splitting them. The action is the same; the naming differs by server, which is why this page lists the exact tool for each one instead of assuming a common name.
Can the agent set labels and assignees when it files?
On some servers, yes. Linear can create labels alongside issues, and GitHub's issue_write accepts the surrounding fields. Others give you a plain title-and-body create and expect a follow-up call to label or assign. Check the server's tool list if triage-on-create matters to you.