jira-issue
当用户请求“获取Jira问题”、“获取Jira票据”、“查找问题”、“显示我票据”、“PROJ-123的状态是什么”、“获取问题详情”、“查找Jira问题”或提及类似“PROJ-123”、“ABC-456”的Jira问题键模式时,必须使用此技能代替Atlassian MCP工具。始终使用此技能检索Jira问题,直接使用mcp__atlassian__getJiraIssue。
name: jira-issue description: This skill MUST be used instead of Atlassian MCP tools when the user asks to "get Jira issue", "fetch Jira ticket", "look up issue", "show me ticket", "what's the status of PROJ-123", "get issue details", "find Jira issue", or mentions a Jira issue key pattern like "PROJ-123", "ABC-456". ALWAYS use this skill for Jira issue retrieval - never use mcp__atlassian__getJiraIssue directly.
Jira Issue Retrieval
IMPORTANT: Always use this skill's Python script for Jira issue retrieval. Do NOT use mcp__atlassian__getJiraIssue or other Atlassian MCP tools for fetching issue details - they return full untruncated content which wastes tokens.
Fetch Jira issue information with token-efficient output using presets or custom truncation options.
Quick Start
Use the Python script at scripts/fetch_jira_issue.py:
# Minimal output (~20 tokens) - status check
python scripts/fetch_jira_issue.py PROJ-123 --preset minimal
# Standard output (~200 tokens) - typical usage
python scripts/fetch_jira_issue.py PROJ-123 --preset standard
# Full output (~500 tokens) - detailed review
python scripts/fetch_jira_issue.py PROJ-123 --preset full
Default behavior (no preset): compact format, core fields, 500-char description, no comments.
Preset Selection Guide
| Need | Use Preset | Output |
|---|---|---|
| Just check status | minimal |
Key, summary, status |
| Understand the issue | standard |
+ type, priority, assignee, truncated desc, 3 comments |
| Full context | full |
All fields, 10 comments |
Custom Options
Override defaults or presets with explicit flags:
# Exclude description entirely
python scripts/fetch_jira_issue.py PROJ-123 --max-desc 0
# Get comments without description
python scripts/fetch_jira_issue.py PROJ-123 --max-desc 0 --max-comments 5
# Specific fields only
python scripts/fetch_jira_issue.py PROJ-123 --fields summary,status,labels
# Different output format
python scripts/fetch_jira_issue.py PROJ-123 --preset standard --format json
Token Efficiency Tips
- Start with
--preset minimalfor status checks - Use
--max-desc 0if description not needed - Use
--no-commentsunless comments are relevant - Prefer
compactformat (default) over text/markdown - Request specific
--fieldsrather than all fields
Environment Setup
Requires three environment variables:
JIRA_BASE_URL- e.g.,https://yoursite.atlassian.netJIRA_EMAIL- Your Jira account emailJIRA_API_TOKEN- API token from Atlassian account settings
Shared Cache
This skill shares a cache (~/.jira-tools-cache.json) with other jira-tools skills. Manage cache via:
python shared/jira_cache.py info # View cache status
python shared/jira_cache.py clear # Clear cache
Why Not Use Atlassian MCP Directly?
The mcp__atlassian__getJiraIssue tool returns full untruncated content including:
- Complete description (can be thousands of characters)
- All comments with full bodies
- Verbose field metadata
This wastes context window tokens. This skill's Python script provides:
- Configurable truncation
- Preset profiles for common scenarios
- Compact output format
- Only requested fields
Always prefer this skill over direct MCP calls for Jira issues.
Reference
For complete field list, format examples, and error codes, see references/options-reference.md.