Agent skill

work-items

Optimize and analyze work items for LLM-assisted development

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/development/work-items

SKILL.md

Work Items Skill

Apply these patterns when working with issues, tickets, or work items from any system (GitHub Issues, Jira, Azure DevOps, Linear, etc.).

LLM Optimization Patterns

File Path Compression

Use glob patterns to compress multiple file paths:

Before After
routes/dev-routes.bicep, routes/test-routes.bicep, routes/prod-routes.bicep routes/{dev,test,prod}-routes.bicep
src/components/Button.tsx, src/components/Input.tsx src/components/{Button,Input}.tsx

Terse Language Rules

  • No articles ("the", "a") unless needed for clarity
  • No filler phrases ("In order to", "This will")
  • Use arrows for transformations: external -> internal, HTTP -> HTTPS
  • Use equals for settings: ingress.external = false

Output Structure

Optimize work items into this token-efficient format:

markdown
# {ID}: {title}

## Context
{1-2 sentence summary}

## Files
- {file paths with glob patterns}

## Changes
1. {file}: {action}

## Verify
- {non-obvious verification step}

Acceptance Criteria Filtering

Remove obvious/implicit criteria:

  • "Deployment succeeds"
  • "No downtime"
  • "Tests pass"
  • "Code reviewed"
  • "PR merged"

Keep non-obvious verification:

  • Specific HTTP status codes
  • Portal/UI status checks
  • Security constraints
  • Integration behaviors

Task Breakdown Rules

Granularity

  • 1 task = 1 logical change that can be committed independently
  • Each task: ~5-30 min of work
  • Group related file changes if they must be atomic

Sequencing

  1. Infrastructure/Config first (env vars, config files, dependencies)
  2. Core changes second (main logic, modules)
  3. Dependent files third (routes, integrations, consumers)
  4. Verification last (manual checks, cleanup)

Task Format

markdown
## Task {n}: {brief title}

**Files**: {file(s) to modify}

**Do**:
- {specific action}

**Test**:
- {how to verify}

Completeness Criteria

Title

  • Descriptive (> 5 words typically)
  • Clearly indicates what feature/change is needed
  • Avoids vague terms: "fix bug", "update code", "improvements"

Description

  • Exists and is not empty
  • Explains what needs to be done
  • Explains why it's needed (business value/context)
  • Provides enough context for a developer to understand scope

Acceptance Criteria

  • Present and populated
  • Contains specific, testable conditions
  • Uses clear language (Given/When/Then or bullet points)
  • Defines what "done" looks like

Didn't find tool you were looking for?

Be as detailed as possible for better results