smithery.ai

Delegating to Jira Agent

Recognize Jira queries and delegate to specialized sub-agent to avoid context pollution

Installation

$ npx skills add https://smithery.ai

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery.ai · top by installs.

npx skills add https://smithery.ai

Browse all from smithery.ai

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Skill metadata

Parsed from SKILL.md frontmatter.

Version1.0.0

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,830 B
  • docs SUMMARY.md 119 B

History

  1. First recorded snapshot · 1 installs

SKILL.md

Delegating to Jira Agent

Core Principle

Never handle Jira operations directly. Always delegate to a specialized sub-agent to keep your context clean and costs low.

Recognition Patterns

Delegate when user says:

  • "show me jira issue PROJ-1234"
  • "what jira issues do I have?"
  • "search jira for nginx"
  • "create a jira for this work"
  • "add a comment to PROJ-1234"
  • Any mention of: jira, issue, ticket, PROJ-, AcmeED-

How to Delegate

Use the Task tool with a specialized prompt:

Task(
  subagent_type: "general-purpose",
  description: "Query Acme Jira",
  prompt: "<full agent instructions from AGENT-INSTRUCTIONS.md>"
)

Agent Prompt Template

When delegating, include:

  1. The complete agent instructions (see AGENT-INSTRUCTIONS.md)
  2. The user's specific request
  3. Clear output format requirements

Example:

You are a Acme Jira specialist. Your job is to query Acme's Atlassian Jira using shell wrappers and return clean results.

<AGENT INSTRUCTIONS HERE>

USER REQUEST: Show me my jira issues for production-main

Return a clean summary with:
- Issue keys
- Summaries
- Status
- URLs

After Agent Returns

  1. Present results cleanly to user
  2. Offer follow-up if relevant (e.g., "Would you like details on PROJ-1234?")
  3. Don't expose mechanics (curl, auth, etc.) to user

Benefits

  • ✅ Main context stays clean
  • ✅ Cheaper queries (sub-agent uses less expensive model)
  • ✅ Specialized knowledge isolated
  • ✅ Scalable pattern for other services

Example Flow

User: "show me my open jira issues"

Main Assistant: [Recognizes Jira query]
              → Invokes Task tool with agent instructions
              → Agent runs issuetracker-mine wrapper
              → Agent returns formatted results

Main Assistant: "You have 20 open issues. Here are the highlights:
                - PROJ-1638: production-main (In Development)
                - PROJ-2216: Reduce catdv log noise (Scheduled)
                ..."

Red Flags

DON'T:

  • ❌ Try to run issuetracker-* scripts yourself
  • ❌ Construct curl commands in main session
  • ❌ Load detailed Jira API knowledge
  • ❌ Handle authentication directly

DO:

  • ✅ Immediately delegate on Jira keywords
  • ✅ Trust the sub-agent's results
  • ✅ Present clean summaries to user

Version History

  • 1.0.0 (2025-10-14): Initial delegation skill created to reduce context pollution