Connecting AI Assistants to Jira and Confluence

Overview

The AMRIT Atlassian MCP Server enables AI assistants to securely access Jira and Confluence using your own credentials.

Once connected, the AI assistant can:

  • Search Jira issues

  • Review project status

  • Summarize tickets

  • Identify blockers

  • Search Confluence documentation

  • Generate meeting summaries

  • Draft status updates

  • Retrieve project information using natural language

Prerequisites

Before connecting your AI assistant, ensure you have:

  • Access to Jira

  • Access to Confluence

  • Personal access tokens for Jira and Confluence

  • An MCP-compatible AI client (Claude Desktop/ Claude Code)

  • Node.js (LTS version recommended) https://nodejs.org/en/download

Generate Your Personal Access Tokens

Jira

  1. Log in to Jira.

  2. Open your account settings.

  3. Navigate to API Tokens / Personal Access Tokens.

  4. Create a new token.

  5. Copy and securely store the token.

Confluence

  1. Log in to Confluence.

  2. Open your account settings.

  3. Navigate to Personal Access Tokens.

  4. Create a new token.

  5. Copy and securely store the token.

Important: Treat these tokens like passwords. Do not share them with anyone.

MCP Endpoint

Use the following AMRIT Atlassian MCP endpoint:

https://amrittools.piramalswasthya.org/atlassian/mcp

Configuration

Finding the Claude Desktop Configuration File

To configure MCP servers, you need to edit the Claude Desktop configuration file.

Windows

Open File Explorer and navigate to:

%APPDATA%\Claude\claude_desktop_config.json (Win+R)

Or directly:

C:\Users\<YourUsername>\AppData\Roaming\Claude\claude_desktop_config.json

macOS

Open Finder and navigate to:

~/Library/Application Support/Claude/claude_desktop_config.json

Linux

Navigate to:

~/.config/Claude/claude_desktop_config.json

Opening the Configuration File

You can open the file using any text editor such as:

  • Notepad (Windows)

  • TextEdit (macOS)

  • VS Code

  • Sublime Text

  • Nano/Vim (Linux)


Add the following MCP configuration to your AI client. 

{
  "mcpServers": {
    "mcp-atlassian": {
      "type": "http",
      "url": "https://amrittools.piramalswasthya.org/atlassian/mcp",
      "headers": {
        "X-Atlassian-Jira-Url": "https://support.piramalfoundation.org/jira",
        "X-Atlassian-Jira-Personal-Token": "<YOUR_JIRA_TOKEN>",
        "X-Atlassian-Confluence-Url": "https://pmp.piramalswasthya.org/confluence",
        "X-Atlassian-Confluence-Personal-Token": "<YOUR_CONFLUENCE_TOKEN>"
      }
    }
  }
}

Replace:

  • <YOUR_JIRA_TOKEN>

  • <YOUR_CONFLUENCE_TOKEN>

with your own credentials(which you just created from Jira and Confluence).

After updating the configuration, completely close(this is very important,close from task managers) and reopen Claude Desktop for the changes to take effect.


Jira Capabilities Available Through the Atlassian MCP Server

The Atlassian MCP Server provides access to Jira project data, issue tracking, agile boards, service management, reporting, and collaboration features through natural language interactions.

Jira Issues

Used for managing Jira work items.

Capabilities include:

  • View issue details

  • Create issues

  • Update issues

  • Assign issues

  • Change issue status

  • Manage priorities

  • Manage labels and components

  • Manage epics, stories, tasks, and bugs

  • Track issue lifecycle

  • Search issues across projects

Typical use cases:

  • Reviewing assigned work

  • Updating ticket status

  • Creating defects and enhancement requests

  • Sprint execution tracking

Jira Search & Fields

Used for discovering and querying Jira metadata.

Capabilities include:

  • Search issues using Jira fields

  • View custom fields

  • Retrieve issue types

  • Retrieve priorities

  • Retrieve workflow statuses

  • Discover project-specific metadata

  • Filter issues based on business criteria

Typical use cases:

  • Reporting

  • Advanced issue searches

  • Data analysis

  • Project audits

Jira Agile

Used for Agile project management.

Capabilities include:

  • Access Scrum boards

  • Access Kanban boards

  • View active sprints

  • View sprint backlog

  • Review sprint progress

  • Track sprint velocity

  • Monitor sprint completion

Typical use cases:

  • Sprint planning

  • Daily stand-ups

  • Sprint reviews

  • Release planning

Jira Links & Versions

Used for dependency and release management.

Capabilities include:

  • View linked issues

  • Track blockers and dependencies

  • View parent-child relationships

  • Manage releases

  • View version progress

  • Monitor release readiness

  • Track release scope

Typical use cases:

  • Dependency management

  • Release coordination

  • Impact analysis

  • Project planning

Jira Comments & Worklogs

Used for collaboration and effort tracking.

Capabilities include:

  • View issue discussions

  • Add comments

  • Summarize discussions

  • View worklogs

  • Analyze effort spent

  • Generate work summaries

  • Track resource utilization

Typical use cases:

  • Team collaboration

  • Project reporting

  • Effort analysis

  • Stakeholder communication

Jira Attachments

Used for managing files associated with issues.

Capabilities include:

  • View attachments

  • Upload attachments

  • Access supporting documents

  • Review screenshots

  • Review design documents

  • Review implementation artifacts

Typical use cases:

  • Bug investigation

  • Documentation review

  • Design validation

  • Knowledge sharing

Jira Service Management

Used for support and service desk operations.

Capabilities include:

  • View service requests

  • Manage customer tickets

  • Track request status

  • Monitor SLAs

  • Review escalations

  • Analyze support workload

Typical use cases:

  • Helpdesk operations

  • Incident management

  • Customer support

  • SLA monitoring

Jira Forms & Metrics

Used for reporting and project analytics.

Capabilities include:

  • Generate project metrics

  • Calculate sprint velocity

  • Analyze cycle times

  • Analyze resolution times

  • Generate workload reports

  • Generate project health reports

  • Identify bottlenecks

  • Create executive summaries

Typical use cases:

  • Project governance

  • Leadership reporting

  • Performance monitoring

  • Delivery tracking

Confluence Capabilities Available Through the Atlassian MCP Server

The Atlassian MCP Server provides access to Confluence knowledge bases, project documentation, attachments, comments, and enterprise search capabilities through natural language interactions.

Confluence Pages

Used for creating, reading, updating, and managing Confluence pages.

Capabilities include:

  • View page content

  • Create new pages

  • Update existing pages

  • Retrieve page metadata

  • View page hierarchy

  • Browse spaces

  • Access page history

  • Review page versions

  • Organize documentation structures

Typical use cases:

  • Reading project documentation

  • Creating meeting notes

  • Maintaining SOPs

  • Updating project status pages

  • Managing knowledge repositories

Confluence Search

Used for discovering information across Confluence spaces.

Capabilities include:

  • Search pages by keyword

  • Search across spaces

  • Locate project documentation

  • Find technical documents

  • Find meeting notes

  • Discover related content

  • Search by title or content

  • Retrieve relevant documentation quickly

Typical use cases:

  • Finding implementation documents

  • Locating project specifications

  • Discovering onboarding materials

  • Retrieving historical decisions

  • Performing knowledge discovery

Confluence Attachments

Used for managing files stored within Confluence pages.

Capabilities include:

  • View page attachments

  • Access uploaded documents

  • Retrieve supporting files

  • Review PDFs and spreadsheets

  • Download attachment metadata

  • Upload supporting documentation

  • Manage project artifacts

Typical use cases:

  • Reviewing design documents

  • Accessing requirement specifications

  • Downloading reports

  • Sharing project deliverables

  • Managing supporting documentation

Confluence Comments

Used for collaboration and discussion around documentation.

Capabilities include:

  • View page comments

  • Add comments

  • Review discussion history

  • Summarize conversations

  • Track feedback

  • Capture review discussions

  • Identify action items from comments

Typical use cases:

  • Document reviews

  • Stakeholder feedback collection

  • Requirement clarification

  • Project discussions

  • Knowledge-sharing conversations

Common Business Use Cases

Business Analysts (BSA)

  • Search requirement documents

  • Retrieve historical decisions

  • Review project specifications

  • Create documentation drafts

  • Analyze stakeholder feedback

Project Managers

  • Review project documentation

  • Track project updates

  • Summarize meeting notes

  • Prepare status reports

  • Monitor project knowledge repositories

Scrum Masters

  • Access sprint documentation

  • Review retrospective notes

  • Track process documentation

  • Share team knowledge

Product Owners

  • Review requirements

  • Analyze feature documentation

  • Gather stakeholder inputs

  • Maintain product knowledge bases

Technical Teams

  • Access architecture documents

  • Review implementation guides

  • Retrieve deployment procedures

  • Search technical knowledge repositories

Security Model

All Confluence operations are performed using the user's own credentials.

Users can only access:

  • Spaces they are authorized to view

  • Pages they have permission to access

  • Attachments available to their account

  • Comments visible within their permissions

The MCP server does not grant additional permissions beyond those already assigned within Confluence.

Example Prompts

After connecting successfully, you can ask:

Jira

  • Search for all issues in the 'PROJ' project where the status is 'In Progress' but the issue hasn't been updated in the last 7 days. For each issue found, use the `jira_add_comment` tool to add a polite comment asking the assignee for a status update. If the issue is a priority 'High' or 'Highest', also transition the issue to 'Blocked'.
  • I need to break down Epic 'PROJ-100' (User Authentication). First, get the details of the Epic. Then, use the `jira_batch_create_issues` tool to create the following 3 tasks in the 'PROJ' project:
    1. "Design auth flow UI" (Assign to UI/UX component)
    2. "Implement JWT token validation" (Backend component)
    3. "Write integration tests for auth endpoints" (QA component)
    Make sure to link all of them to Epic 'PROJ-100' using the additional_fields parameter.
  • Find the 'Engineering' scrum board using `jira_get_agile_boards`. Retrieve the currently 'active' sprint from it. Then, fetch all issues in that sprint. Give me a structured summary of:
    1. What was completed yesterday (look at recent transitions to Done).
    2. What is currently 'In Progress'.
    3. Any issues that are flagged or 'Blocked'.
  • Create a new sprint on board ID 42 named "Sprint 25 - Q3 Core Features" starting today and ending in 14 days. Once created, search the backlog using JQL for the top 5 highest priority unresolved issues in project 'DEV' with the label 'q3-roadmap'. Add these issues to the newly created sprint using the `jira_add_issues_to_sprint` tool.
  • Get the issues from Queue ID 10 in Service Desk ID 4. For each issue, use the `jira_get_issue_sla` tool to check the 'time_in_status' and 'resolution_time' metrics. Identify any issues that are close to breaching their SLA. Add an internal comment (restricted to the 'jira-administrators' group) to those issues warning the team, and transition them to 'Escalated'.
  • Retrieve the ProForma form details attached to issue 'SUP-404' using `jira_get_issue_forms` and `jira_get_form_details`. Read the answers provided by the customer. If the customer indicated that their system is completely down, update the Jira issue priority to 'Highest' and add a worklog of '15m' for triage time.
  • Fetch the development info for issue 'API-505' using `jira_get_issue_development_info`. Check if there is a merged pull request attached. If there is:
    1. Transition the issue to 'Done'.
    2. Use `jira_create_version` to ensure version 'v1.4.0' exists in the 'API' project.
    3. Update the issue to set its fixVersion to 'v1.4.0'.
    4. Leave a comment thanking the developer and logging '1h' of time spent.
  • Search for all Epics in project 'CORE' that were transitioned to 'Done' this week. For each Epic, use the `jira_create_remote_issue_link` tool to attach a URL pointing to our Confluence design space (https://confluence.example.com/core-designs), titled "Final Design Documentation", with the relationship set to "documentation".
  • Use `jira_get_issue_images` to view the images attached to bug ticket 'WEB-99'. Analyze the UI error shown in the screenshot. Then, use `jira_update_issue` to append a detailed technical description of the visual error to the issue's existing description. Include what the expected behavior should be based on standard UI/UX patterns.
  • Run `jira_batch_get_changelogs` for issues 'PROJ-1', 'PROJ-2', and 'PROJ-3'. Filter the changelogs specifically for the 'status' and 'assignee' fields over the last 5 days. Generate a timeline report showing exactly who changed the statuses and when, and highlight any instances where an issue bounced back and forth between 'In Review' and 'In Progress'.

Confluence

  • Create a new Confluence page in the 'ENG' space titled "Weekly Engineering Sync - [Insert Today's Date]". Use the page creation tool to set the parent page to ID '10293847'. The page content should use standard Atlassian storage format (HTML) with the following structure:
    1. An "Attendees" section with a bulleted list.
    2. An "Agenda" section.
    3. A "Discussion" section.
    4. An "Action Items" section using task list formatting.
    Once created, return the URL of the new page.
  • Retrieve the current content of the Confluence page "ADR-014: Database Selection" (Page ID: 55667788) using the page retrieval tool. Read the existing content. Then, use the page update tool to append a new section at the very bottom titled "Review - Q3 2026". In this section, write a brief summary stating that the decision to use PostgreSQL is still valid and performing well under load. Ensure you increment the page version number correctly to successfully save the update.
  • Use the Confluence Search tool to run a CQL (Confluence Query Language) query: `space = "DEV" AND lastModified < now("-52w")`. This will find all pages in the DEV space that haven't been updated in over a year. For each page in the results, fetch the page details to identify the creator. Then, use the comments tool to add a comment to each page saying: "Automated Ping: This page hasn't been updated in over a year. Could you please review it for accuracy and update or archive it?"
  • Search Confluence using the search tool for pages containing the exact phrase "Environment Setup" or "Local Dev" across both the 'Engineering' and 'Operations' spaces. Retrieve the top 5 most relevant results. Read the content of those 5 pages, synthesize the prerequisite installation steps (like Docker, Node, etc.), and provide me with a consolidated, step-by-step markdown summary with links back to the original source pages.
  • Fetch all comments on the Confluence page "Q4 Payment Gateway Integration" (Page ID: 998877). Analyze the comment threads and categorize them into two buckets: "Resolved/Agreed" and "Open Questions". Present the open questions to me as a bulleted list, noting who asked the question and what part of the document they are referencing.
  • I have reviewed the draft for the "API Rate Limiting" page (ID: 112233). Use the comments tool to add a new top-level comment to this page. The comment should read: "Great start! However, we need to explicitly define the HTTP 429 response headers we plan to return (e.g., Retry-After). Let's add a table for that in the implementation section."
  • Check the attachments for the Confluence page "Release v2.4.0" (Page ID: 445566). Use the attachment listing tool to see if a file named 'architecture-diagram-v2.png' is already present. If it is not, use the attachment upload tool to attach the local file located at `./assets/architecture-diagram-v2.png` to the page. Include a comment on the upload: "Finalized architecture diagram for v2.4.0 release."
  • Use the Confluence search tool to find the page titled "Legacy Auth Flow Diagram". Once found, get the page ID and use the attachment retrieval tool to download the latest version of the file named "oauth-flow.pdf". After retrieving the file, analyze its contents and generate a text-based plantUML sequence diagram that represents the flow shown in the PDF, so we can embed it directly into the page text.

Security Considerations

  • Never share personal access tokens.

  • Tokens should be stored only in your AI client configuration.

  • Access is limited to resources your account is already authorized to view.

  • The MCP server does not grant additional permissions.

Troubleshooting

Authentication Errors

Verify:

  • The Jira token is valid.

  • The Confluence token is valid.

  • Tokens have not expired.

  • Tokens were copied correctly.

Unable to Access Tickets or Pages

Verify:

  • You have access to the requested project or space.

  • The requested resource exists.

  • Your account permissions are sufficient.

Connection Issues

Verify:

  • The MCP endpoint URL is correct.

  • Your network connection is available.

  • Your AI client supports MCP connections.

Support

For issues related to MCP connectivity or access, contact the AMRIT Agentic AI Framework team and provide:

  • AI client being used

  • Error message received

  • Jira project or Confluence space involved

  • Steps performed before the issue occurred

  • No labels