Skip to content

About

AI-powered Jira development workflow for coding agents — manage Jira issues, create Git branches, link development work, and track sprints directly from your IDE.

Topics

Resources

Contributing

Security policy

Stars

2 stars

Watchers

0 watching

Forks

Latest commit

 

History

27 Commits

Folders and files

Repository files navigation

Jira Development Workflow — Agent Plugin

A task-oriented Jira development workflow plugin that sits between an AI coding agent and the Jira/Git tools it needs for day-to-day delivery work. It packages a structured workflow for reading issues, planning implementation, starting work, creating branches, linking development artifacts back to Jira, and checking sprint progress — without requiring raw API calls from the user.

This plugin is workflow guidance on top of the official Atlassian Rovo MCP v2 server. It does not replace Atlassian's MCP product; it teaches a coding agent a safe Jira Cloud delivery loop (read, plan, start work, branch, link, sprint, complete).

Built on the open Agent Plugins 1.0.0 specification — works with any compatible client.

Breaking in 2.0.0: Jira access uses Atlassian's hosted Rovo MCP v2 (https://mcp.atlassian.com/v2/mcp) with client-managed OAuth. This plugin supports Atlassian Cloud only. Self-hosted Jira and API tokens in plugin environment variables are no longer part of the package.

Compatible Clients

Client Install Method
Kiro (IDE/CLI/Crew) Powers panel → Add Custom Power → Import from GitHub
VS Code Agent Plugins setup
Cursor Plugins docs
GitHub Copilot About plugins
ChatGPT / Codex OpenAI plugins

Note: Each client discovers plugin.json, loads skills from skills/, and starts MCP servers from mcp.json — but the activation UX (keyword triggers, manual enable) varies per client.

Quick Start

1. Install

From GitHub (any compatible client):

https://github.com/ShreyPurohit/Jira-Dev-Workflow-Plugin

Point your client's "Add Plugin" / "Import from GitHub" flow at this URL.

Or clone locally:

git clone https://github.com/ShreyPurohit/Jira-Dev-Workflow-Plugin.git

Then import the local directory in your client.

2. Connect to Atlassian

In your compatible MCP/Agent Plugin client, add or select the official Atlassian Rovo MCP v2 connection:

https://mcp.atlassian.com/v2/mcp

Use the client's Atlassian authorization flow to authenticate the connection. OAuth 2.1 is the recommended method for interactive use. The plugin does not launch a local MCP process. Per Agent Plugins, mcp.json has no portable credential fields — do not put tokens or Authorization headers in the plugin.

Official setup: Getting started with the Atlassian Rovo MCP Server.

3. Verify

Open your AI client and try:

"Show me the details of PROJ-123"

If you get issue details back, the plugin is working.

What You Can Do

Prompt What happens
"Summarize PROJ-123" Fetches and presents the issue with status, description, and ACs
"What are the acceptance criteria for PROJ-123?" Extracts and lists ACs from the description
"Plan PROJ-123" Creates a structured implementation plan from the issue requirements
"Start work on PROJ-123" Moves the Jira issue into the appropriate In Progress status
"Create a branch for PROJ-123" Generates a Jira-aware Git branch name and creates the branch
"Link my work to PROJ-123" Posts a structured Jira comment linking branch/commit/PR context
"What's my sprint looking like?" Shows current sprint tickets, priorities, and blocker status
"What's blocking?" Surfaces blocked or stalled sprint items
"Move PROJ-123 to In Review" Safely transitions the issue using dynamic discovery
"Comment on PROJ-123" Adds a structured progress or completion comment
"Complete PROJ-123" Verifies current state and transitions the issue through completion
"Show me open tasks in PROJ" Searches with appropriate JQL
"Log 2h on PROJ-123" Records a worklog (time-tracking) entry against the issue
"Prepare PROJ-123 for review" Generates a PR/MR description from ACs + git diff, offers transition
"Download the images from PROJ-123" Fetches issue/comment attachments via the authenticated MCP server

Plugin Structure

jira-dev-workflow-plugin/
├── plugin.json                 # Agent Plugin manifest (identity + keywords)
├── mcp.json                    # MCP server configuration
├── README.md                   # This file
├── LICENSE                     # MIT license
├── PRIVACY.md                  # Privacy policy
├── assets/logo.svg             # Plugin logo
├── .gitignore
├── skills/
│   ├── jira-read/SKILL.md          # Read/search/summarize issues
│   ├── jira-plan/SKILL.md          # Create implementation plans
│   ├── jira-start-work/SKILL.md    # Move a ticket into an in-progress status
│   ├── jira-update-status/SKILL.md # Transition issues safely
│   ├── jira-comment/SKILL.md       # Add structured comments
│   ├── jira-complete/SKILL.md      # Complete development workflows
│   ├── jira-branch/SKILL.md        # Create a Jira-aware Git branch
│   ├── jira-link-work/SKILL.md     # Link Git work back to Jira
│   ├── jira-sprint/SKILL.md        # Show sprint status and blockers
│   ├── jira-worklog/SKILL.md       # Log time (worklog) against an issue
│   ├── jira-review-prep/SKILL.md   # Generate a PR description + review transition
│   └── jira-attachments/SKILL.md   # Download issue/comment attachments

How it works

  1. plugin.json declares the plugin identity, keywords, and version following the Agent Plugins 1.0.0 spec.
  2. mcp.json configures the official Atlassian Rovo MCP v2 server over Streamable HTTP (type + url only). The hosted endpoint is https://mcp.atlassian.com/v2/mcp. Authentication is managed by the compatible MCP client.
  3. Skills are Agent Skills (SKILL.md) that teach the agent how to use Rovo MCP Jira tools, including resolving cloudId via getAccessibleAtlassianResources before other calls.
  4. When you mention a Jira issue key or ask about a Jira task, the appropriate skill activates and guides the agent through the correct workflow.

Design principles

  • Task-oriented, not API-oriented. Skills represent what developers ask ("start work on X") rather than raw tool names.
  • Safety by default. All write operations require user confirmation.
  • Dynamic discovery. Transitions are discovered at runtime, not hardcoded — works with any Jira workflow.
  • Jira and Git remain separate capabilities. Starting work and creating a branch are composable but independent operations.
  • Composition over duplication. Some skills are task-oriented shortcuts over shared primitives (jira-start-work / jira-complete / jira-review-prep over jira-update-status; jira-plan consumes jira-attachments) — not competing implementations of the same mechanic.
  • No invented data. The agent only reports what Jira or Git returns; it never guesses.
  • Portable. Built on open standards (Agent Plugins + Agent Skills + MCP) — not locked to any single IDE.

Skills

Skill Purpose
jira-read Read, summarize, and search Jira issues
jira-plan Analyze requirements and create implementation plans
jira-start-work Move a Jira issue into the appropriate In Progress status
jira-update-status Safely transition Jira issues between statuses
jira-comment Add structured progress or completion comments to Jira
jira-complete Complete Jira work with a completion comment and status update
jira-branch Create a Jira-aware Git branch from the issue summary
jira-link-work Link Git work artifacts back to Jira
jira-sprint Show sprint progress, assignee workload, and blockers
jira-worklog Log time spent (worklog entries) against a Jira issue
jira-review-prep Generate a PR/MR description from ACs + git diff, then review
jira-attachments Download issue/comment attachments via the authenticated MCP

Jira ↔ Git Workflow

These capabilities are intentionally separate, but they can be composed when the user asks for both.

Start work

jira-start-work
      ↓
Jira status → In Progress

This skill reads the ticket, finds the correct transition dynamically, confirms before mutating Jira, and verifies the result. It does not create or inspect a Git branch.

Create branch

jira-branch
      ↓
Git branch → feat/PROJ-123-...

This skill reads the issue summary, generates a Jira-aware branch name, checks repository safety, asks for confirmation, and creates the branch. It does not transition the Jira issue to In Progress.

Link development work

jira-link-work
      ↓
Git work → Jira

This skill connects branch, commit, or PR context back to the Jira issue using a structured comment.

Configuration Reference

This plugin does not take Jira credentials or project-filter environment variables. Agent Plugins mcp.json only declares the remote server:

{
  "$schema": "https://agent-plugins.org/schemas/1.0.0/mcp.schema.json",
  "mcpServers": {
    "atlassian-rovo-mcp": {
      "type": "streamable-http",
      "url": "https://mcp.atlassian.com/v2/mcp"
    }
  }
}

To limit searches to a project, name the project in the prompt. Skills encode that as JQL (project = KEY). If several Atlassian Cloud sites are available, the agent asks which site to use after getAccessibleAtlassianResources.

MCP Servers

This plugin uses the official Atlassian Rovo MCP v2 server:

Server Transport Endpoint Purpose
atlassian-rovo-mcp streamable-http https://mcp.atlassian.com/v2/mcp Atlassian Jira access

The server is hosted by Atlassian. The plugin does not launch a local MCP process. Authentication is handled by the compatible MCP client. Some Rovo MCP calls consume Rovo credits on the Atlassian site.

Safety Conventions

Operation Safety level
Read issue / search ✅ No confirmation needed
Get transitions (read-only) ✅ No confirmation needed
Transition issue ⚠️ Confirms before executing
Add comment ⚠️ Shows preview, confirms before posting
Edit issue fields ⚠️ Confirms before executing
Delete / close 🛑 Explicit confirmation required

Transition safety

The most critical safety rule: always match transitions by destination status, never by transition label.

Jira transition names are arbitrary workflow labels — the agent discovers transitions dynamically and matches your intent against where the transition actually goes.

Troubleshooting

"Jira authentication failed"

Reconnect or re-authorize the Atlassian Rovo MCP v2 connection in your compatible MCP/Agent Plugin client. Confirm the client completed OAuth for the correct Atlassian Cloud site. If the problem persists, verify that the authorized Jira account has access to the relevant projects and operations.

"Missing cloudId" / tools fail after login

The first Jira call must be getAccessibleAtlassianResources. Later tools need that site's cloudId. If multiple sites are listed, choose one explicitly.

"Issue not found"

  • Check the issue key is correct (e.g., PROJ-123, not proj-123)
  • Verify you have access to the project in Jira
  • Confirm the selected Atlassian Cloud site is the one that contains the issue

"No transition available"

  • The issue is in a status that doesn't allow the requested transition
  • Use "Show me the transitions for PROJ-123" to see what's available
  • Your Jira workflow may require intermediate steps

Version Notes

The repository is currently on version 2.1.1 as defined in plugin.json. Version 2.1.1 builds on the twelve-skill workflow introduced in v2.1.0, improving Jira attachment discovery, requirement-relevant file handling, and download failure guidance.

  • Jira-aware branch creation
  • Git-to-Jira linking work
  • Sprint visibility and blocker reporting
  • Worklog / time-tracking entries (jira-worklog)
  • Review preparation — PR/MR description from ACs + git diff (jira-review-prep)
  • Authenticated attachment downloads (jira-attachments), also consumed by jira-plan
  • Client-managed OAuth against Atlassian Cloud

Release notes for each version are documented in CHANGELOG.md.

Support

For bugs, questions, or feature requests, open a GitHub issue or email the address above. Do not include Jira API tokens, credentials, or private ticket data in public issues.

Privacy

This plugin is open-source and does not operate a hosted backend. Privacy policy: PRIVACY.md.

Security

If you discover a security vulnerability, please do not disclose it publicly through GitHub Issues. Follow the responsible disclosure process described in SECURITY.md.

Never include Jira API tokens, credentials, or other sensitive information in public issues or reports.

Contributing

We welcome contributions! For guidelines on development setup, testing, and submitting changes, see CONTRIBUTING.md.

Key points:

  • Connect and authorize the official Atlassian Rovo MCP v2 server through your client (Atlassian Cloud)
  • Test in your Agent Plugin client
  • Follow the skill design principles (single responsibility, no overlap, safety first)
  • Keep plugin.json / mcp.json within the closed Agent Plugins 1.0.0 schemas
  • Update documentation and CHANGELOG.md for release-relevant changes

Standards

License

MIT — see LICENSE.

About

AI-powered Jira development workflow for coding agents — manage Jira issues, create Git branches, link development work, and track sprints directly from your IDE.

Topics

Resources

Contributing

Security policy

Stars

2 stars

Watchers

0 watching

Forks

Releases

Contributors