atlassian-mcp-server
This MCP server enables LLM agents to interact with Atlassian Jira and Confluence Cloud through comprehensive tools for managing issues, pages, and content.
Jira Capabilities:
Search & List Issues: Use JQL queries to find and list issues with customizable result limits
Issue Management: View full issue details, create new issues with project/type specifications, and update issue summaries/descriptions
Comments: Add comments (supports plain text or Atlassian Document Format) and retrieve existing comments
Workflow: Transition issues between statuses using transition IDs and discover available transitions for specific issues
Attachments: Download files attached to Jira issues
Confluence Capabilities:
Page Management: List pages within spaces, view page content/metadata, create new pages (optionally nested under parent pages), and delete pages
Content Editing: Update page content with automatic version handling and support for Mermaid diagrams via code blocks
Search: Perform advanced searches using CQL with customizable result limits and criteria like labels and titles
Comments: Retrieve all comments on Confluence pages
Special Features:
Supports Atlassian Document Format (ADF) for rich text in Jira descriptions and comments
Automatic version management for Confluence page updates with manual control option
Provides tools for interacting with Atlassian Cloud services, including Jira issue management (search, create, update, comment, transition) and Confluence page management (search, view, create, edit, delete).
Allows searching and listing pages within spaces, viewing page content, creating new pages (including nested pages), editing pages with auto-versioning, and deleting pages using CQL search capabilities.
Enables searching issues with JQL, viewing issue details, creating and updating issues with rich text support, adding comments, and transitioning issues through workflows.
Click on "Deploy 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., "@atlassian-mcp-server list Jira issues assigned to me that are overdue"
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.
Atlassian MCP Server
An MCP server for Jira and Confluence Cloud. Write in markdown — the server converts to Atlassian Document Format (ADF) automatically.
Quickstart
git clone https://github.com/tingyiy/atlassian-mcp-server.git
cd atlassian-mcp-server
pip install -r requirements.txt # or: uv pip install -r requirements.txtThen add it to your MCP client (see Configuration below).
You'll need an Atlassian API token and these values:
Variable | Example |
|
|
| your API token |
|
|
|
|
|
|
Related MCP server: Jira
Tools
Jira
Tool | Description |
| Search issues using JQL |
| Get full issue details |
| Create an issue (markdown description) |
| Update summary or description (markdown) |
| Add a comment (markdown) |
| Retrieve all comments |
| Download an attachment by ID |
| Move issue through workflow |
| List available transitions |
Confluence
Tool | Description |
| List pages in a space |
| Get page content and metadata |
| Create a page (optionally nested) |
| Update a page (auto-increments version) |
| Delete a page |
| Search using CQL |
| Get all comments on a page |
| Add a comment (with reply support) |
| Download an image attachment |
Markdown Support
Jira tools (jira_create_issue, jira_update_issue, jira_add_comment) accept markdown for descriptions and comments. The server converts it to ADF behind the scenes using md2adf.
Supported formatting: headings, bold, italic, strikethrough, inline code, links, images, code blocks (with language), bullet/ordered/nested lists, blockquotes, horizontal rules, and tables.
Configuration
Claude Code
Add to ~/.claude.json:
{
"mcpServers": {
"atlassian": {
"type": "stdio",
"command": "python3",
"args": ["/path/to/atlassian-mcp-server/server.py"],
"env": {
"ATLASSIAN_USERNAME": "you@company.com",
"ATLASSIAN_API_KEY": "your_api_token",
"JIRA_URL": "https://your-domain.atlassian.net/rest/api/3",
"CONFLUENCE_URL": "https://your-domain.atlassian.net/wiki",
"CONFLUENCE_SPACE_KEY": "ENG"
}
}
}
}Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"atlassian": {
"command": "python3",
"args": ["/path/to/atlassian-mcp-server/server.py"],
"env": {
"ATLASSIAN_USERNAME": "you@company.com",
"ATLASSIAN_API_KEY": "your_api_token",
"JIRA_URL": "https://your-domain.atlassian.net/rest/api/3",
"CONFLUENCE_URL": "https://your-domain.atlassian.net/wiki",
"CONFLUENCE_SPACE_KEY": "ENG"
}
}
}
}Cursor / VS Code
Add to .cursor/mcp.json or .vscode/mcp.json:
{
"mcpServers": {
"atlassian": {
"command": "python3",
"args": ["/path/to/atlassian-mcp-server/server.py"],
"env": {
"ATLASSIAN_USERNAME": "you@company.com",
"ATLASSIAN_API_KEY": "your_api_token",
"JIRA_URL": "https://your-domain.atlassian.net/rest/api/3",
"CONFLUENCE_URL": "https://your-domain.atlassian.net/wiki"
}
}
}
}Make sure the python3 in command points to an environment with the dependencies installed. To be explicit, use the full path (e.g. /path/to/venv/bin/python).
Notes
Mermaid Diagrams in Confluence
Confluence's Mermaid plugin stores diagram source internally, so the MCP tools can't render them programmatically. Instead, provide the Mermaid source in a code block — the user can then convert it in the Confluence editor.
License
MIT — see LICENSE.
Available Tools
22 toolsconfluence_add_commentB
Adds a comment to a Confluence page. Set parent_comment_id to reply to an existing comment.
Inputs: page_id (required), body (required), parent_comment_id. Returns: result
confluence_create_pageC
Creates a new Confluence page, optionally under a parent page.
Inputs: title (required), content (required), parent_id, space_key. Returns: result
confluence_delete_pageC
Deletes a Confluence page.
Inputs: page_id (required). Returns: result
confluence_get_attachment_imageC
Gets an image attachment on a Confluence page and returns it as an Image.
Inputs: page_id (required), filename (required).
confluence_get_commentsC
Gets all comments for a Confluence page.
Inputs: page_id (required). Returns: result
confluence_searchB
Searches Confluence content using CQL (Confluence Query Language). Example: title ~ "meeting" AND label = "notes"
Inputs: cql (required), limit. Returns: result
edit_confluence_pageB
Updates a Confluence page. If version is not provided, it will be automatically incremented.
MERMAID DIAGRAMS: Confluence Cloud uses the Mermaid Diagrams plugin. You CANNOT create rendered diagrams programmatically. The mermaid-cloud macro only references diagram content in the plugin's internal storage (not accessible via API).
To include a Mermaid diagram, provide it as a code block for the user to manually convert: <ac:structured-macro ac:name="code" ac:schema-version="1"> <ac:parameter ac:name="language">text</ac:parameter> ac:plain-text-body</ac:plain-text-body> </ac:structured-macro>
The user can then convert this code block to a rendered diagram in the Confluence editor.
Inputs: page_id (required), title (required), content (required), version. Returns: result
jira_add_commentA
Adds a comment to a Jira issue.
Args: issue_key: The ID or key of the issue. comment: The comment content in markdown. Supports headings, bold, italic, strikethrough, links, code blocks, lists, tables, etc.
Mentioning users: To tag a user, first call jira_search_users to find their accountId, then use @[accountId] in the text (e.g. @[712020:abc123]). You can also use @username which auto-resolves if there is exactly one match, but will fail if ambiguous — prefer @[accountId].
Inputs: issue_key (required), comment (required). Returns: result
jira_create_issueA
Creates a new Jira issue. For description, accepts a markdown string. Supports headings, bold, italic, strikethrough, links, code blocks, lists, tables, etc.
Mentioning users: To tag a user, first call jira_search_users to find their accountId, then use @[accountId] in the text (e.g. @[712020:abc123]). You can also use @username which auto-resolves if there is exactly one match, but will fail if ambiguous — prefer @[accountId].
Inputs: project_key (required), summary (required), description, issuetype. Returns: result
jira_delete_commentA
Deletes a comment from a Jira issue.
Args: issue_key: The ID or key of the issue. comment_id: The ID of the comment to delete (from jira_get_comments).
Inputs: issue_key (required), comment_id (required). Returns: result
jira_download_attachmentB
Downloads a Jira attachment by ID, saves it to a temp file, and returns the file path.
Inputs: attachment_id (required). Returns: result
jira_edit_commentA
Edits an existing comment on a Jira issue.
Args: issue_key: The ID or key of the issue. comment_id: The ID of the comment to edit (from jira_get_comments). comment: The new comment content in markdown. Supports headings, bold, italic, strikethrough, links, code blocks, lists, tables, etc.
Mentioning users: To tag a user, first call jira_search_users to find their accountId, then use @[accountId] in the text (e.g. @[712020:abc123]). You can also use @username which auto-resolves if there is exactly one match, but will fail if ambiguous -- prefer @[accountId].
Inputs: issue_key (required), comment_id (required), comment (required). Returns: result
jira_get_commentsB
Gets all comments for a Jira issue.
Inputs: issue_key (required). Returns: result
jira_get_transitionsC
Gets available transitions for a Jira issue.
Inputs: issue_key (required). Returns: result
jira_list_fieldsA
Lists all available Jira fields (standard and custom).
Use this to discover field IDs you can pass as additional_fields to read_jira_issue, list_jira_issues, or jira_update_issue.
Inputs: no named parameters. Returns: result
jira_search_usersA
Searches for Jira users by name or email.
Args: query: Name, email, or username fragment to search for. max_results: Maximum number of results to return.
Inputs: query (required), max_results. Returns: result
jira_transition_issueA
Transitions a Jira issue to a new status using a transition ID. Use jira_get_transitions to find available transition IDs.
Inputs: issue_key (required), transition_id (required). Returns: result
jira_update_issueA
Updates fields on a Jira issue.
Args:
issue_key: The issue key (e.g. "PROJ-123").
summary: New summary text.
description: New description in markdown. Supports headings, bold,
italic, strikethrough, links, code blocks, lists, tables, etc.
additional_fields: Dict of field_id -> value for simple fields
(plain text, dates, numbers, selects).
Use jira_list_fields to discover field IDs.
Example: {"duedate": "2026-04-01", "labels": ["bug"]}
rich_text_fields: Dict of field_id -> markdown string for rich text
fields that need ADF conversion (e.g. Developer Notes,
Deployment Info). These get converted to Atlassian Document
Format automatically.
Example: {"customfield_10402": "## Plan
Step 1
Step 2"}
Mentioning users (in description and rich_text_fields): First call jira_search_users to find the accountId, then use @[accountId] in the text (e.g. @[712020:abc123]).
Inputs: issue_key (required), summary, description, additional_fields, +1 more. Returns: result
list_confluence_pagesC
Lists Confluence pages in a space.
Inputs: space_key, limit. Returns: result
list_jira_issuesA
Lists Jira issues using JQL.
By default returns key, summary, status, priority, assignee. Use additional_fields to include extra Jira fields in the response.
Args: jql: JQL query string. next_page_token: Token for pagination (returned in previous response). max_results: Maximum number of results to return. additional_fields: Extra Jira field names to include, e.g. ["issuetype", "labels", "created", "updated", "duedate", "components", "fixVersions", "reporter", "resolution", "parent", "subtasks"].
Inputs: jql, next_page_token, max_results, additional_fields. Returns: result
read_jira_issueA
Gets details of a specific Jira issue.
By default returns a compact set of fields. Use additional_fields to request more without bloating every response.
Args: issue_key: The issue key (e.g. "PROJ-123"). additional_fields: Extra Jira field names to include, e.g. ["components", "fixVersions", "duedate", "resolution", "parent", "subtasks", "issuetype", "customfield_XXXXX"].
Inputs: issue_key (required), additional_fields. Returns: result
view_confluence_pageC
Gets the content of a Confluence page.
Inputs: page_id (required). Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 22 tools
Tools are clearly separated by product prefix (confluence_ vs jira_) and have distinct actions on different resources (pages, issues, comments). No functional overlap between tools.
Most tools follow a verb_noun pattern with product prefixes (e.g., confluence_create_page, jira_add_comment). Minor inconsistencies: 'edit_confluence_page' vs 'update', and 'list_' versus '_list' patterns (e.g., list_confluence_pages vs jira_list_fields).
22 tools split across two products (Confluence ~7, Jira ~15) is well-scoped. Each tool serves a clear purpose without excessive overlap or missing essential operations.
Core CRUD operations for Confluence pages and Jira issues are present, along with comments, search, transitions, and user lookup. Minor gaps like attachment upload are missing, but the surface covers primary workflows.
Maintenance
Related MCP Connectors
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
The Remote MCP server acts as a standardized bridge between LLM applications (like Claude, ChatGPT, and Cursor) and external services, enabling AI agents to access external tools and resources. Its primary capability is providing a centralized search tool to discover other MCP servers and their respective tools. Unlike local implementations, it runs remotely with OAuth authentication and permission controls for security.
An MCP server that provides an API to LLMs to manage their JumpCloud resources.
Official Fibery MCP server: query and update your workspace from any AI client via OAuth.
Related MCP Servers
- AlicenseBqualityDmaintenanceProvides capabilities for searching Jira issues using JQL and retrieving detailed issue information.228MIT
- MIT
- AlicenseNot gradedqualityDmaintenanceA model context server that provides prompts that can be used as slash commands for clients like Zed Editor, in order to add page contents as context to the AI assistant.11MIT
- MIT