Skip to content

Connect OpenClaw to Jira

Let an OpenClaw agent create, search, update and transition issues in Jira, with no credentials handed to the agent.

Tags:integrations

Jira is the issue tracker for engineering and product teams, especially those running sprints, boards and formal workflows. It carries a lot of process, and process needs feeding: bug reports arrive as chat messages and never become tickets, issues sit in “In Progress” long after the work shipped, and the sprint report is only accurate after someone spends an hour fixing states before the retro.

With Claworc’s External Connections, an OpenClaw agent can work inside your Jira site on your behalf. You authorize Atlassian once through a popup, Composio holds the resulting token, and the agent gets an auto-generated claworc-jira skill that describes which Jira tools exist and how to call them. The agent never sees a password, an OAuth token or an API token.

Composio’s Jira toolkit exposes 94 actions covering issues, comments, sprints, boards, projects, users and permissions. In practice an OpenClaw agent leans on a handful of capabilities:

  • Search with JQL — search issues with JQL, which lets the agent express almost any question about the backlog as a query.
  • Read an issue — a single issue with its fields, description, comments and history.
  • Create issues in bulk — create one issue, or up to fifty in a single call when a document turns into a batch of tickets.
  • Edit and assign — update fields such as priority, labels and fix version, and assign the issue to someone.
  • Transition workflow states — see which workflow moves are allowed from the current state and make one of them.
  • Comment on issues — so the reasoning behind a change stays with the ticket.
  • List sprints — the sprints on a board and their state.

Every action goes through a small broker running next to the agent, so the agent only needs to describe what it wants done in plain language and let the skill handle the rest.

Forward a support message or a chat thread and ask the agent to file it. It searches for an existing issue with the same symptoms first, and if there is none, creates a new bug with a clear title, reproduction steps pulled from the report, the right component and priority, and a comment linking back to the source. If it finds a duplicate, it comments on that one instead.

Ask the agent, each evening, to search for issues in “In Progress” whose linked pull request has merged, and move them to “In Review” or “Done” with a comment explaining why. It checks the allowed transitions first, so it never tries a move the workflow forbids. The board reflects reality without anyone dragging cards around.

“What is still open in the current sprint, grouped by assignee, and which of those have no update in five days?” The agent lists the active sprint, runs the JQL, reads the recent comments and returns a short table. The same question works for the last three sprints when you are preparing a retrospective.

Connecting Jira takes three steps: an admin adds a Composio API key to Claworc’s settings, you open the instance’s Settings tab, click + Add connection, pick Jira and sign in to Atlassian. The claworc-jira skill appears in the instance a moment later. The full walkthrough, including the exact Composio permissions the key needs, is in the External Connections documentation.