Jira

@jira show me all open bugs assigned to me in the PLATFORM project, grouped by priority

Found 12 open bugs in PLATFORM assigned to you: Critical (2): • PLATFORM-847: Database connection pool exhausted under load • PLATFORM-923: Auth token refresh fails for OAuth users High (5): • PLATFORM-891: API rate limit headers missing in responses • PLATFORM-902: Webhook delivery retries not respecting backoff • PLATFORM-915: Memory leak in background job processor • PLATFORM-931: Cache invalidation race condition on deploys • PLATFORM-944: Error logs not capturing full stack traces Medium (5): • PLATFORM-856: Dashboard charts load slowly with large datasets • PLATFORM-878: Export CSV button disabled for non-admin users • PLATFORM-889: Tooltip positioning breaks on mobile viewport • PLATFORM-912: Notification emails use wrong timezone • PLATFORM-938: Search autocomplete shows archived projects

This example demonstrates the MCP's search and filtering capabilities. Requires read:jira-work OAuth scope. The grouping and formatting are handled by the AI based on raw issue data returned by Jira's API—actual response structure depends on your project's custom fields and workflow states.

@jira create a bug ticket in MOBILE project: login screen crashes on iOS 16 when biometric auth fails, assign to sarah@company.com, priority high

Created MOBILE-1847: Login screen crashes on iOS 16 when biometric auth fails Type: Bug Priority: High Assignee: Sarah Chen (sarah@company.com) Status: To Do Reporter: You The issue has been added to the MOBILE project backlog. Sarah will receive a notification about the assignment. You can view the full ticket at https://yourcompany.atlassian.net/browse/MOBILE-1847

This showcases the Create Issue tool with assignee lookup by email. Requires write:jira-work scope. The MCP resolves the email to a Jira account ID automatically—if the email isn't found in your Jira instance, creation will fail with a clear error. Side-effect: creates a permanent ticket and triggers assignee notifications.

@jira read the last 15 comments on BACKEND-456 and summarize what's blocking the release

Based on 15 comments spanning the last 8 days, three blockers are preventing BACKEND-456 from releasing: 1. **Database migration rollback plan** (comments from Alex, Maria): The schema changes in PR #2891 don't have a tested rollback procedure. QA flagged this on Jan 8th; no update since. 2. **Performance regression in staging** (comments from Dev team): Load tests show 40% slower response times compared to v2.3. Jordan suspects the new caching layer; profiling in progress. 3. **Documentation incomplete** (comments from Product): API endpoint changes aren't reflected in the public docs. Technical writing team needs 2-3 days once the endpoint signatures are finalized. Most recent activity suggests the performance issue is the critical path—other blockers are waiting on that resolution.

This example pairs Jira's comment-fetching with AI synthesis to extract signal from discussion threads. Requires read:jira-work scope. The AI interprets comment content and timestamps—accuracy depends on how clearly your team documents blockers in comments. No side-effects; this is a read-only operation.

When Jira MCP handles support-to-eng handoffs at 20+ tickets/day

A 6-person support team fields 30-40 customer issues daily and escalates 8-10 to engineering as Jira bugs. The Jira MCP wins here because OAuth2 keeps credentials out of shared prompts, and Bulk Create Issues lets the AI draft multiple tickets from a single support thread without rate-limit drama. The team uses Add Comment to append customer context and Assign Issue to route by component owner. The trade-off: if your support tool already has a native Jira integration (Zendesk, Intercom), that's faster for one-off escalations. But when you're triaging a batch of related issues in Switchy—say, five customers hit the same edge case—the MCP's bulk operations and attachment uploads make it the right call. Worth it if you escalate more than 15 tickets a week.

Why this MCP scales sprint prep across time zones

A 9-person product team spans three time zones and runs two-week sprints with 40-60 stories per cycle. The Jira MCP is the move when async sprint planning happens in Switchy: the AI can Create Sprint, then use Bulk Create Issues to draft the backlog from a product brief or customer research doc. Add Watcher to Issue keeps stakeholders looped without manual tagging, and the 46-tool scope means you can also query sprint velocity or reassign mid-sprint without leaving the workspace. The boundary: if your team is under 5 people and everyone's in the same standup, the overhead of OAuth setup and learning the MCP's tool names isn't worth it—just use Jira's UI. But once you're coordinating across Slack threads and Google Docs with 8+ contributors, the MCP's batch operations and comment threading save 2-3 hours per sprint cycle.

When Jira MCP turns incident notes into actionable tasks

An 11-person engineering team runs a post-incident review after a 90-minute outage and needs to convert the retrospective doc into 12 follow-up tasks across three Jira projects. The Jira MCP is built for this: the AI reads the incident timeline, uses Bulk Create Issues to generate tasks with correct project keys and priorities, then Add Attachment to link the runbook PDF. Create Issue handles the parent epic, and Assign Issue routes remediation work to the right on-call rotation. The limit: if your incidents are rare (under one per quarter) or your retro process is highly templated, the MCP's 46-tool surface area is overkill—a Zapier action or manual entry is faster. But for teams shipping daily with 2+ incidents per month, the MCP's ability to parse unstructured notes and create structured work in one pass justifies the OAuth setup. Buy if you're documenting more than 8 action items per incident.