J-Assist by Optmyzr

51 tools · MCP live Sign in

MODEL CONTEXT PROTOCOL Now in public beta

Talk to Jira the way
you talk to Claude.

J-Assist is the Optmyzr MCP server that connects Claude to Jira. Search tickets, transition statuses, log work, attach files — in plain English, with your permissions and a full audit trail.

Read the docs

SSO · Google Workspace/ Optmyzr accounts only/ No credit card

J-Assist · one prompt, two interfaces — drag to compare

Claude.ai Terminal

claude.ai · web claude code · terminal

← drag to reveal · same prompt, both surfaces →

// How it works

Three steps. Then just type.

J-Assist is a Model Context Protocol server. It exposes Jira to Claude as 51 typed, scoped actions — including bulk operations and Product Discovery (JPD) insights — not a hand-rolled prompt and not a screen-scraper.

STEP 01 / SIGN IN

Sign in with Google.

Authenticate with your Optmyzr Google Workspace account. J-Assist provisions a workspace, picks up your team, and respects SSO.

# nothing to install. browser only. $ open j-assist.optmyzr.com ✓ SSO via @optmyzr.com ✓ workspace ready

STEP 02 / CONNECT JIRA

Authorize via Jira OAuth.

One click. J-Assist redirects you to Atlassian, you approve scoped read/write access, you're back — 51 tools provisioned against your instance. Add more instances anytime.

# /connect/jira → redirect id.atlassian.com/authorize ✓ scopes: read:jira-work · write:jira-work ✓ jira.atlassian.net · 51 tools live

STEP 03 / CONNECT CLAUDE

Paste the URL into Claude.

In Claude.ai or Claude Desktop, open Settings → Connectors → Add custom and paste the J-Assist URL. That's it. On Claude Code (Windows / Linux / macOS), one curl wires it up. See the Claude Code guide →

# claude.ai → connectors → add custom "https://j-assist.optmyzr.com/mcp" # claude code · one-liner $ curl https://j-assist.optmyzr.com/install | sh

// 51 tools — including bulk & JPD

Every Jira move Claude can make, with a name.

Each tool is a typed JSON-RPC call — predictable, scoped to your permissions, and logged. Group below to filter.

jira_get_boardread

Get a single Jira board by ID, or list boards filtered by project, type, or name.

jira_get_changelogread

Status change history and computed metrics — cycle time, lead time — for tickets.

jira_get_create_metaread

List required and allowed fields for creating an issue in a project / issue type.

jira_get_myselfread

Get the current authenticated user's info including account ID.

jira_get_ticketread

Fetch full details of a Jira ticket — description, comments, attachments, links.

jira_get_worklogsread

Get time-logged entries for a Jira ticket, with author and duration.

jira_list_instancesread

List all configured Jira instances with their names, URLs, and project prefixes.

jira_list_projectsread

List all accessible projects on a Jira instance.

jira_list_sprintsread

List sprints on a Jira board, filterable by state — active, future, closed.

jira_searchread

Search Jira tickets using JQL — full power of the Jira query language.

jira_search_usersread

Search for Jira users by name or email. Returns account IDs for assignment.

jira_download_attachmentread

Download a specific file attachment from a Jira ticket by filename.

jira_get_componentsread

List all components for a project — Backend, Frontend, API, etc.

jira_get_prioritiesread

Get all available priority levels — Blocker, Critical, Major, Minor, Trivial.

jira_get_sprintread

Get sprint details — name, goal, state, start/end dates, and origin board.

jira_get_sprint_issuesread

Get all issues in a sprint with status, assignee, priority. Supports JQL filtering within the sprint.

jira_list_link_typesread

List available issue link types — Blocks, Duplicates, Relates, Causes — with inward/outward names.

jira_create_ticketwrite

Create a new Jira ticket in a project with type, priority, assignee, labels, parent.

jira_update_ticketwrite

Update fields on an existing Jira ticket. Validates field types and required values.

jira_transitionwrite

Change the status of a Jira ticket, auto-handling intermediate workflow steps.

jira_link_issueswrite

Create a link — Blocks, Duplicate, Relates — between two Jira issues.

jira_delete_linkdelete

Remove a link between two Jira issues.

jira_move_to_epicwrite

Move one or more issues into an epic. Preserves status, assignee, and links.

jira_delete_ticketdelete

Permanently delete a Jira ticket. Optionally deletes subtasks too.

jira_add_remote_linkwrite

Attach a web URL — PRs, docs, Figma, Confluence — to a Jira ticket.

jira_create_sprintwrite

Create a new sprint on a scrum board. Starts in ‘future’ state, ready to be planned and started.

jira_update_sprintwrite

Update a sprint's name, goal, or dates. Partial update — only provided fields change.

jira_start_sprintwrite

Activate a future sprint with start and end dates. Only one sprint can be active per board.

jira_complete_sprintwrite

Close an active sprint. Incomplete issues stay — move them to the next sprint manually.

jira_bulk_get_ticketsbulk

Fetch up to 100 tickets in a single round trip using POST /issue/bulkfetch.

jira_bulk_create_ticketsbulk

Create many tickets at once (50/request, auto-chunked). Optional inline attachments per ticket. Great for QA seeding test data.

jira_bulk_update_ticketsbulk

Apply the same field changes — labels, priority, fix versions — to N tickets in parallel.

jira_bulk_assignbulk

Reassign many tickets to one user. Accepts a display name (auto-resolves to accountId) or accountId directly.

jira_bulk_add_commentbulk

Post the same comment on N tickets in parallel — useful for “deployed to staging, please verify” announcements.

jira_bulk_link_issuesbulk

Link many source issues to one target with the same link type. Common pattern: “block these 5 tickets with PR-123.”

jira_bulk_transitionbulk

Transition many tickets to the same status in parallel. Sprint cleanup in one call.

jira_move_to_sprintbulk

Move up to 50 issues into a sprint in a single call with optional ranking. This IS the bulk sprint-move.

jira_add_commentwrite

Add a comment to a Jira ticket, with optional inline file attachments.

jira_edit_commentwrite

Replace the text of an existing comment on a Jira ticket.

jira_reply_commentwrite

Reply to a specific comment on a Jira ticket, quoting the original.

jira_delete_commentdelete

Permanently delete a comment from a Jira ticket.

jira_log_workwrite

Log time spent on a Jira ticket with an optional comment.

jira_add_attachmentwrite

Attach a small file (≤1 MiB) via inline base64. For larger files use jira_request_upload_url.

jira_request_upload_urlwrite

Get a one-time signed PUT URL. The client streams the file directly to Jira — bytes never traverse the LLM context window. Up to 50 MiB.

jira_refresh_cachewrite

Force-refresh cached field metadata for an instance. Use after a workflow change.

jira_add_watcherwrite

Add a user as a watcher to receive notifications for all updates on an issue.

jira_delete_attachmentdelete

Permanently delete an attachment from a Jira ticket by its attachment ID.

jpd_get_insight_countjpd

Read the insight count for a JPD idea via the standard Jira API. Reliable, fast, read-only.

jpd_list_insightsjpd

List insights attached to a JPD idea — description, source URLs, snippets — via the Polaris GraphQL API.

jpd_create_insightjpd

Add a new insight to a JPD idea with description and optional source URL. Experimental Polaris mutation.

jpd_describe_typejpd

Introspect the live Polaris GraphQL schema for any type. Use to discover real field names when the experimental API drifts.

// Trust model

Your data. Your scopes. Your audit trail.

J-Assist runs as an MCP server in Optmyzr's hardened cloud. We're a passthrough — not a copy.

Google Workspace SSO only.

Sign-in via your @optmyzr.com Google account. No passwords, no shared secrets, no API keys floating in a desktop config.

Scoped OAuth, not write-everything tokens.

Per-instance scopes are visible at install. Connect read-only first, expand to write when you trust it. Revoke from one panel.

Per-action audit trail.

Every tool call is logged with the caller, the arguments, the result, and the duration. Export to your SIEM via webhook.

Dry-run for any write.

Bulk transitions, label rewrites, batch comments — preview the diff, confirm once, execute. Roll back from the audit log.

Stop tab-hopping.
Just ask Claude.

Sign in with your Optmyzr Google account. Connect Jira in 30 seconds. Have your first conversation with your backlog today.

Optmyzr accounts · Free during public beta

Talk to Jira the wayyou talk to Claude.