Serveur MCP Atlassian pour JIRA

VérifiéSûr

Le serveur MCP Atlassian permet aux agents de codage comme Cursor et Claude Code d'interagir directement avec JIRA et Confluence. Il fournit des outils pour consulter, créer, mettre à jour et rechercher des tickets JIRA, ainsi que pour gérer des commentaires et des pages Confluence. Utile pour les développeurs qui ont besoin d'accéder aux données Atlassian sans quitter leur environnement de codage.

Spar Skills Guide Bot
DeveloppementIntermédiaire
6002/06/2026
Claude CodeCursor
#atlassian#jira#mcp#issue-tracking#confluence

Recommandé pour

Notre avis

Ce skill permet d'interagir avec JIRA et Confluence depuis des agents de codage via le serveur MCP officiel d'Atlassian.

Points forts

  • Accès direct aux issues JIRA et pages Confluence depuis l'IDE.
  • Opérations courantes (création, modification, transition) facilement scriptables.
  • Recherche avancée avec JQL intégrée.
  • Configuration centralisée via le fichier mcp.json.

Limites

  • Nécessite une authentification OAuth préalable avec Atlassian.
  • Dépend de la disponibilité du serveur MCP externe.
  • Fonctionnalités Confluence limitées par rapport à l'API complète.
Quand l'utiliser

Utilisez ce skill lorsque vous avez besoin de consulter ou modifier des issues JIRA directement depuis votre environnement de développement.

Quand l'éviter

Évitez de l'utiliser pour des opérations de masse ou lors de l'absence de connexion internet, car le serveur est externe.

Analyse de sécurité

Sûr
Score qualité88/100

The skill documents setup and usage of an official Atlassian MCP server for JIRA and Confluence interactions. No destructive commands, remote code execution, or data exfiltration instructions are present. All operations are performed through a controlled API, and authentication is handled via OAuth.

Aucun point d'attention détecté

Exemples

Get JIRA issue details
Get details of JIRA issue PROJ-123 using the Atlassian MCP server.
Search issues with JQL
Search for JIRA issues in project PROJ with component 'backend' and status 'In Progress' using JQL.
Create a new issue
Create a new JIRA issue in project PROJ with issue type 'Story', summary 'Implement login feature', and component 'auth'.

name: atlassian-mcp description: Use the Atlassian MCP Server to interact with JIRA from coding agents (Cursor, Claude Code). This skill documents how to use the official Atlassian MCP server for JIRA and Confluence operations. Enable this MCP server in your Cursor/Claude Code settings for JIRA access. requires:

  • atlassian-oauth

Atlassian MCP Server for JIRA

The Atlassian MCP Server provides direct access to JIRA and Confluence from coding agents. This is the supported way to access Atlassian data from Cursor or Claude Code.

Setup

In Cursor

The Atlassian MCP Server is already configured globally. To enable it:

  1. Open Cursor Settings
  2. Go to Tools & MCP
  3. Enable "Atlassian-MCP-Server"

Configuration location: ~/.cursor/mcp.json

{
  "mcpServers": {
    "Atlassian-MCP-Server": {
      "url": "https://mcp.atlassian.com/v1/sse"
    }
  }
}

In Claude Code

Add to your Claude Code MCP settings:

{
  "mcpServers": {
    "Atlassian-MCP-Server": {
      "url": "https://mcp.atlassian.com/v1/sse"
    }
  }
}

Available Tools

When enabled, the Atlassian MCP Server provides these tools:

Issue Operations

| Tool | Description | | ---------------------------------------- | ------------------------------- | | mcp_Atlassian_getJiraIssue | Get details of a specific issue | | mcp_Atlassian_createJiraIssue | Create a new issue | | mcp_Atlassian_editJiraIssue | Update an existing issue | | mcp_Atlassian_transitionJiraIssue | Change issue status | | mcp_Atlassian_searchJiraIssuesUsingJql | Search issues with JQL |

Comments

| Tool | Description | | ------------------------------------- | -------------------------- | | mcp_Atlassian_addCommentToJiraIssue | Add a comment to an issue | | mcp_Atlassian_getJiraIssueComments | Get comments from an issue |

Confluence (if enabled)

| Tool | Description | | ------------------------------------ | ------------------------ | | mcp_Atlassian_createConfluencePage | Create a Confluence page | | mcp_Atlassian_getConfluencePage | Get a Confluence page | | mcp_Atlassian_searchConfluence | Search Confluence |

Common Patterns

Get Issue Details

Use mcp_Atlassian_getJiraIssue with issueKey: "PROJ-123"

Search for Issues

Use mcp_Atlassian_searchJiraIssuesUsingJql with:
  jql: "project = YOUR_PROJECT AND component = 'YOUR_COMPONENT' AND status = 'In Progress'"

Create an Issue

Use mcp_Atlassian_createJiraIssue with:
  project: "YOUR_PROJECT"
  issueType: "Story"
  summary: "Implement new feature"
  description: "Description of the feature"
  components: ["YOUR_COMPONENT"]

Transition Issue Status

Use mcp_Atlassian_transitionJiraIssue with:
  issueKey: "PROJ-123"
  transition: "In Progress"

Add Comment

Use mcp_Atlassian_addCommentToJiraIssue with:
  issueKey: "PROJ-123"
  comment: "Started working on this issue"

JQL Quick Reference

Common JQL queries for the Orient Task Force:

# All issues for YOUR_COMPONENT
project = YOUR_PROJECT AND component = "YOUR_COMPONENT"

# In Progress issues
project = YOUR_PROJECT AND component = "YOUR_COMPONENT" AND status = "In Progress"

# Blockers
project = YOUR_PROJECT AND component = "YOUR_COMPONENT" AND (priority = Blocker OR labels = blocked)

# Recent issues (last 7 days)
project = YOUR_PROJECT AND component = "YOUR_COMPONENT" AND created >= -7d

# My issues
project = YOUR_PROJECT AND component = "YOUR_COMPONENT" AND assignee = currentUser()

# Sprint issues
project = YOUR_PROJECT AND component = "YOUR_COMPONENT" AND sprint in openSprints()

Troubleshooting

"Tool not found" Error

Make sure the Atlassian MCP Server is enabled in your IDE settings. The tools won't appear until the server is connected.

Authentication Issues

The Atlassian MCP Server uses OAuth. If you're not authenticated:

  1. The tool will prompt you to authenticate
  2. Follow the OAuth flow in your browser
  3. Grant the necessary permissions

Rate Limits

The Atlassian MCP Server is subject to Atlassian's rate limits. If you hit limits:

  • Reduce query frequency
  • Use more specific JQL to reduce result sets
  • Cache results when appropriate
Skills similaires