European-Parliament-MCP-Server: Skill for Claude Code

GitHub Agentic Workflows Skill

Purpose

Comprehensive guidance for GitHub Agentic Workflows (gh-aw) — a Go-based GitHub CLI extension that enables writing agentic workflows in natural language using markdown files, compiled into GitHub Actions workflows with defense-in-depth security. Also covers Copilot coding agent orchestration, MCP server configuration, stacked PRs, and OWASP Agentic security.

When to Use

• ✅ Writing markdown-defined AI workflows (gh-aw) for automated repository tasks
• ✅ Orchestrating Copilot coding agent assignments with base_ref and custom_instructions
• ✅ Building multi-step agentic workflows with stacked PRs
• ✅ Configuring safe outputs, permissions, and AI engine selection
• ✅ Securing agentic pipelines against prompt injection and data exfiltration
• ✅ Implementing Continuous AI patterns (scheduled + event-triggered AI automation)

GitHub Agentic Workflows (gh-aw)

Overview

GitHub Agentic Workflows (gh-aw) augments deterministic CI/CD with Continuous AI — systematic, automated application of AI to software collaboration. Workflows are defined in markdown with YAML frontmatter and compiled to GitHub Actions via gh aw compile.

Key capabilities: Issue triage, CI failure diagnosis, documentation maintenance, code quality improvement, metrics/analytics, security scanning, multi-repo coordination, project planning.

Workflow Definition Format

---
timeout-minutes: 5
on:
  schedule: daily           # or cron: "0 9 * * 1-5", or event triggers
  issue:
    types: [opened, reopened]
permissions:
  contents: read
  issues: read
  pull-requests: read
tools:
  github:
    toolsets: [issues, labels, pull_requests, repos]
safe-outputs:
  create-issue:
    title-prefix: "[report] "
    labels: [automated, report]
    close-older-issues: true
  add-labels:
    allowed: [bug, feature, enhancement, documentation]
  add-comment: {}
  create-pull-request:
    title-prefix: "[auto] "
    max-changed-files: 5
---

# Workflow Title

Natural language instructions for the AI agent describing what to do.

## Context
- Repository-specific guidance
- What to analyze or act on

## Rules
- Constraints and boundaries
- Output format requirements