mcp-jira-scoped
Provides tools to interact with Jira, including reading and searching issues, creating and updating issues, adding comments, managing transitions, linking issues, and handling attachments through Atlassian's scoped API tokens.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-jira-scopedWhat's the status of PROJ-1234?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-jira-scoped
The only MCP server for Jira that works with Atlassian's scoped API tokens.
Every other Jira MCP server uses classic (unscoped) API tokens with basic auth against yoursite.atlassian.net. Atlassian is deprecating those. This server uses scoped tokens with the modern api.atlassian.com gateway — the way Atlassian intends these tokens to be used.
What You Can Do
Once connected, you can ask your AI assistant things like:
"What's the status of PROJ-1234?"
"Search for all open bugs assigned to me in the BACKEND project"
"Create a story in PROJ for the database migration, priority P2, under epic PROJ-100"
"Move PROJ-1234 to In Progress"
"Add a comment to PROJ-1234 saying the fix is deployed to staging"
"Link PROJ-1234 as blocking PROJ-5678"
"Find all issues with 'auth' in the summary updated this week"
Related MCP server: Jira MCP Server
Why This Server
mcp-jira-scoped | Other Jira MCP servers | |
Token type | Scoped (modern, | Classic (being deprecated) |
Auth gateway |
|
|
Scope enforcement | Server-side, before every API call | None — relies on AI self-restraint |
Default mode | Read-only (write must be explicitly granted) | Full access |
Delete safety | Requires | No guard |
Quick Start
1. Create a Scoped API Token
Go to Atlassian API Tokens
Click "Create API token with scopes"
Grant the scopes you need:
read:jira-work— read issues, search, comments, projectswrite:jira-work— create/update/delete issues, add commentsread:jira-user— look up usersread:me— read your own profile
2. Add to Your AI Client
Claude Desktop / Claude Code
Add to your .mcp.json:
{
"mcpServers": {
"jira": {
"command": "npx",
"args": ["-y", "mcp-jira-scoped"],
"env": {
"JIRA_INSTANCE": "yourcompany",
"JIRA_USER_EMAIL": "you@yourcompany.com",
"JIRA_API_TOKEN": "<your-scoped-token>",
"JIRA_SCOPES": "read:jira-work,write:jira-work"
}
}
}
}Cursor
Add the same config to Cursor's MCP settings (Settings > MCP Servers).
VS Code (Copilot)
Add to .vscode/mcp.json:
{
"servers": {
"jira": {
"command": "npx",
"args": ["-y", "mcp-jira-scoped"],
"env": {
"JIRA_INSTANCE": "yourcompany",
"JIRA_USER_EMAIL": "you@yourcompany.com",
"JIRA_API_TOKEN": "<your-scoped-token>",
"JIRA_SCOPES": "read:jira-work,write:jira-work"
}
}
}
}Configuration
Env Var | Required | Description |
| Yes | Instance name (e.g. |
| Yes | Scoped API token ( |
| Yes | Email associated with the token |
| No | Comma-separated scopes. Defaults to |
| No | Atlassian Cloud ID. Auto-fetched if not set. Find it at |
Available Tools (14)
Read Tools (read:jira-work)
Tool | Description |
| Get issue by key. Pass |
| Search issues via JQL with pagination |
| List comments on an issue |
| List accessible projects |
| Get project details by key |
| List available issue link types |
| List all fields (system + custom) — discover |
| List attachments on an issue with filename, size, MIME type, and download URL |
| Get available transitions for an issue with required screen fields expanded |
Write Tools (write:jira-work)
Tool | Description |
| Create an issue (task, bug, story, epic). Supports custom fields |
| Update fields on an issue. Supports custom fields |
| Add a comment (plain text auto-converted to ADF) |
| Move an issue to a new status. Accepts |
| Delete an issue (requires |
| Link two issues (blocks, relates, split, clone, etc.) |
| Upload a local file as an attachment to an issue |
| Delete an attachment by ID (requires |
User Tools (read:jira-user)
Tool | Description |
| Get user info by account ID |
| Search users by name or email |
Safety
Scope enforcement — tools are blocked server-side if their required scope isn't granted. The API call never happens.
Read-only default — if
JIRA_SCOPESis not set, only read tools are even registered.Delete confirmation —
jira_delete_issuerequiresconfirm: true.No token logging — tokens are redacted from all error messages via
sanitizeError.No admin operations — no project creation/deletion, workflow changes, or webhook management. Ever.
Troubleshooting
"Client must be authenticated" (401)
You're probably using a scoped token against the old yoursite.atlassian.net URL. Scoped tokens (ATATT... prefix) only work via api.atlassian.com. This server handles this automatically — make sure you're using mcp-jira-scoped, not another Jira MCP server.
"Failed to parse Connect Session Auth Token" (403)
You're sending a scoped token as a Bearer token. Scoped tokens use Basic auth (email:token) via the api.atlassian.com gateway. Again, this server handles it — this error means you're using a different server.
Token scopes vs server scopes
There are two layers of scope enforcement:
Atlassian's scopes — set when you create the token. These control what Atlassian's API allows.
Server scopes (
JIRA_SCOPESenv var) — control what tools this MCP server makes available. These can only be more restrictive, never less.
If you get a 403 from Atlassian, check that your token has the required scope. If you get a scope enforcement error from the MCP server, check your JIRA_SCOPES env var.
Custom fields
Jira projects often have required custom fields (e.g. "Work Category", "Story Point Estimate"). Use the customFields parameter on jira_create_issue and jira_update_issue:
Create an issue in PROJ with summary "Fix login bug" and set customfield_10016 to 3The AI will pass {"customFields": {"customfield_10016": 3}}. To find custom field IDs, check your Jira project's field configuration.
Cloud ID
The server auto-fetches your Cloud ID from https://yourinstance.atlassian.net/_edge/tenant_info. If this fails (e.g. corporate firewall), set JIRA_CLOUD_ID manually.
Roadmap
See Prioritised-TodoList.md for the full feature checklist — what's built, what's next, and where contributions are welcome. Key areas open for contribution:
Issue tools — assign shortcut, create/transition metadata discovery
Comments — edit and delete
Watchers & Worklogs — full CRUD
Boards & Sprints — Agile workflow support
Attachments — list, download, upload
Development
git clone https://github.com/deepwired/mcp-jira.git
cd mcp-jira
npm install
npm run build
npm testSee CONTRIBUTING.md for details on adding tools and submitting PRs.
Note on Package Naming
The GitHub repo is mcp-jira but the npm package is mcp-jira-scoped. We plan to unify under mcp-jira in a future release. For now, use npx -y mcp-jira-scoped to run the server.
License
Apache 2.0 — see LICENSE.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/deepwired/mcp-jira'
If you have feedback or need assistance with the MCP directory API, please join our Discord server