Skip to content
Back to skills

Run Jira

ASecurity

Fetch a Jira issue and propose an implementation plan based on codebase analysis

  • 3,754 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added September 3, 2026
devopsbashgit

Works with

  • mcp

Security analysis

A100/100

Scanned September 3, 2026

npx -y skills add DataDog/datadog-agent --skill run-jira --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Run Jira?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Run Jira
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/datadog-run-jira/badge)](https://www.skillsdirectory.com/skills/datadog-run-jira)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: run-jira
description: Fetch a Jira issue and propose an implementation plan based on codebase analysis
argument-hint: "<JIRA-KEY e.g. ACTP-1234>"
model: sonnet
---

Fetch the Jira issue **$ARGUMENTS** from the Datadog Atlassian instance and use it as the basis for a codebase analysis and implementation proposal.

## Step 1: Gather Jira issue data

Use the Atlassian MCP tools to fetch the issue. **Only request the fields you need** to avoid huge responses:

1. Call `mcp__atlassian__getJiraIssue` with:
   - `cloudId`: `datadoghq.atlassian.net`
   - `issueIdOrKey`: `$ARGUMENTS`
   - `fields`: `["summary", "description", "status", "assignee", "issuetype", "comment", "priority"]`
2. Extract the **title** (summary), **description**, **status**, **assignee**, and **comments** from the response.

If the issue cannot be found, stop and inform the user.

## Step 2: Summarize the issue

Present a clear summary of the Jira issue:
- **Key**: $ARGUMENTS
- **Title**: the issue summary
- **Status**: current status
- **Assignee**: who is assigned
- **Description**: the full description
- **Comments**: any relevant context or discussion from comments
- **Link**: https://datadoghq.atlassian.net/browse/$ARGUMENTS

## Step 3: Fetch linked resources

Scan the issue description and comments for links to external resources and fetch them for additional context:

- **Datadog notebooks / postmortems** (`app.datadoghq.com/notebook/<id>`): use `mcp__datadog-mcp__get_datadog_notebook` with the notebook ID
- **GitHub PRs** (`github.com/.../pull/<number>`): use `gh pr view <number>` via Bash
- **GitLab commits/pipelines** or other URLs: use `WebFetch` if accessible

This step is critical — linked resources often contain the root cause analysis, timelines, and technical details that the Jira description alone does not capture.

## Step 4: Analyze the codebase

Based on the issue requirements and linked resources, explore the codebase to understand:
- Which files and packages are relevant
- Existing patterns and conventions that should be followed
- Dependencies and potential impacts

Use Glob, Grep, and Read tools extensively. For broad exploration, use the Task tool with `subagent_type=Explore`.

## Step 5: Propose an implementation

Enter plan mode with `EnterPlanMode` and write a detailed implementation plan that includes:
- A breakdown of the changes needed, organized by file
- Any new files that need to be created
- Test strategy
- Potential risks or open questions

Wait for user approval before implementing.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.