Skip to content

README.md

Latest commit

 

History

History
1022 lines (790 loc) · 36.2 KB

README.md

File metadata and controls

1022 lines (790 loc) · 36.2 KB
Raw

🤖 Custom GitHub Copilot Agents - EU Parliament Monitor

This directory contains custom agent profiles for GitHub Copilot, designed to provide domain-specific expertise for the EU Parliament Monitor project.

📋 Overview

Each agent profile is a Markdown file with YAML frontmatter that defines specialized expertise. When working on tasks related to a specific domain, GitHub Copilot can leverage these profiles to provide more informed and contextually appropriate assistance.

Agent Architecture

All agents follow the 2026 GitHub Copilot Coding Agent Standard with:

  • GitHub MCP Insiders API integration for experimental features
  • Toolset policy — omit the tools frontmatter field so that Copilot CLI allows all tools by default. Never write tools: ["*"] inside an mcp-servers block: the gh-aw MCP gateway (awmg) treats * as a literal tool name and exposes 0 tools. See .github/prompts/08-infrastructure.md §2 for the canonical rule.
  • Organization-wide access via PAT token (Hack23 repositories)
  • Modern Copilot features: assign_copilot_to_issue, create_pull_request_with_copilot, stacked PRs, job tracking
  • gh-aw agent imports — workflows can import a single agent file via the imports: frontmatter field (see gh-aw Copilot Custom Agents). In this repository's gh-aw usage (tested against gh-aw v0.69.3, 2026-04-21), imported agent files contribute body-only content: the agent body is appended to the workflow prompt, but agent frontmatter — including mcp-servers: — is not merged into the workflow frontmatter. Therefore mcp-servers: must live in the workflow itself or in a shared frontmatter-only workflow component (e.g. .github/workflows/shared/mcp/news-mcp-servers.md). See .github/agents/news-generation.agent.md for the canonical split-imports pattern used by every article-generating news-*.md workflow.
  • Cross-repository patterns for accessing European-Parliament-MCP-Server, riksdagsmonitor, cia, ISMS-PUBLIC
  • ISMS compliance mapped to ISO 27001, NIST CSF 2.0, CIS Controls v8.1, GDPR, NIS2, EU CRA

🗂️ Role Index (analysis-producer / analysis-consumer / gh-aw-infrastructure)

Group the catalog by their role in the news critical path. See .github/prompts/README.md § Analysis Artifact Integration for the chain they support.

Role Agents
Analysis producers (write artifacts under analysis/daily/<run>/) intelligence-operative, news-generation.agent
Analysis consumers (read artifacts, render to content) news-journalist, marketing-specialist, business-development-specialist
Data / pipeline suppliers (Stage A inputs) data-pipeline-specialist
gh-aw infrastructure (workflow authoring / CI) agentic-workflows.agent, devops-engineer, ci-cleaner.agent, create-safe-output-type.agent, custom-engine-implementation.agent, interactive-agent-designer.agent
Quality / review / docs quality-engineer, grumpy-reviewer.agent, contribution-checker, documentation-architect, technical-doc-writer.agent, w3c-specification-writer.agent
Product / frontend product-task-agent, frontend-specialist, developer.instructions

Canonical analysis anchors every agent on the news critical path should read: ai-driven-analysis-guide.md, artifact-catalog.md, per-artifact-methodologies.md, analysis/templates/README.md, reference-quality-thresholds.json.

🎯 Available Agents

1. 📊 Product Task Agent (product-task-agent)

Expertise: Product management, GitHub issue creation, European Parliament monitoring

When to Use:

  • Creating GitHub issues for new features or improvements
  • Analyzing repository health and quality metrics
  • Coordinating work across multiple specialized agents
  • Prioritizing tasks and managing product backlog
  • European Parliament data integration planning
  • ISMS compliance tracking and auditing

Key Capabilities:

  • Automated issue creation with proper labels and priorities
  • Playwright browser testing for visual regression
  • Multi-language quality assurance (14 languages)
  • WCAG 2.1 AA accessibility auditing
  • European Parliament MCP integration monitoring
  • ISMS policy enforcement (ISO 27001, GDPR, NIS2)
  • Agent coordination and task assignment

Example Use:

@product-task-agent analyze the current state of multi-language support and create issues for any gaps found

2. 📰 News Journalist (news-journalist)

Expertise: The Economist-style European Parliament reporting, multi-language content generation

When to Use:

  • Writing or editing news articles about European Parliament activities
  • Generating week-ahead previews of plenary sessions
  • Creating committee reports and analysis
  • Covering legislative propositions and motions
  • Multi-language content quality assurance
  • SEO optimization and structured data

Key Capabilities:

  • The Economist editorial standards and style
  • European Parliament coverage (MEPs, committees, plenary sessions)
  • Multi-language news generation (14 languages)
  • European Parliament MCP data integration for articles
  • Fact-checking and source verification
  • GDPR-compliant political reporting
  • SEO metadata and structured data generation

Example Use:

@news-journalist create a week-ahead article covering the upcoming plenary session using European Parliament MCP data

3. 🎨 Frontend Specialist (frontend-specialist)

Expertise: HTML5/CSS3, WCAG 2.1 AA accessibility, responsive design, multi-language UI

When to Use:

  • Implementing or improving UI components
  • Ensuring WCAG 2.1 AA accessibility compliance
  • Creating responsive layouts (320px - 1440px+)
  • Multi-language interface design
  • Performance optimization (Core Web Vitals)
  • Security headers configuration (CSP, HSTS)

Key Capabilities:

  • Semantic HTML5 markup
  • CSS3 responsive design patterns
  • WCAG 2.1 AA accessibility implementation
  • Keyboard navigation and screen reader support
  • Multi-language UI patterns (14 languages, RTL support)
  • Core Web Vitals optimization
  • GitHub Pages deployment optimization

Example Use:

@frontend-specialist fix the language switcher to be fully keyboard accessible and WCAG 2.1 AA compliant

4. 🔄 Data Pipeline Specialist (data-pipeline-specialist)

Expertise: European Parliament MCP Server integration, data caching, API client patterns

When to Use:

  • Integrating European Parliament MCP Server data
  • Implementing data caching strategies
  • Building retry logic and error handling
  • Creating fallback mechanisms for MCP unavailability
  • Validating data schemas and quality
  • Optimizing API client performance

Key Capabilities:

  • European Parliament MCP tool mastery (6 tools)
  • src/mcp/ep-mcp-client.ts TypeScript source and gateway setup (scripts/mcp-setup.sh)
  • LRU cache implementation and TTL strategies
  • Exponential backoff retry logic
  • Undici HTTP client patterns
  • Data schema validation
  • Error handling and logging best practices

Example Use:

@data-pipeline-specialist implement caching for MEP data with a 24-hour TTL and retry logic for API failures

5. ⚙️ DevOps Engineer (devops-engineer)

Expertise: GitHub Actions, CI/CD, automation, deployment strategies

When to Use:

  • Creating or modifying GitHub Actions workflows
  • Setting up automated testing pipelines
  • Implementing deployment automation
  • Configuring Node.js and Playwright environments
  • Managing MCP server pre-installation
  • Branch protection and repository settings

Key Capabilities:

  • GitHub Actions workflow authoring
  • Daily news generation automation
  • Playwright browser testing in CI
  • Node.js 25 environment setup
  • MCP server pre-installation and caching
  • GitHub Pages deployment strategies
  • Security scanning integration (Dependabot, CodeQL)

Example Use:

@devops-engineer create a GitHub Actions workflow to validate HTML/CSS on every PR

🔒 Security note — no standalone security-architect agent. Security responsibilities are distributed across every agent via the shared skills catalogue. See .github/skills/README.md § Security Skills and § Compliance Skills for the authoritative entry points: security-by-design, sdlc-security-integration, threat-modeling, isms-compliance, compliance-frameworks, data-protection, and the full Hack23 ISMS-PUBLIC policy → skill map in ISMS_SKILLS_COMPREHENSIVE.md. Security review is an explicit responsibility of grumpy-reviewer.agent and contribution-checker.

6. 📚 Documentation Architect (documentation-architect)

Expertise: C4 models, Mermaid diagrams, API documentation, architecture documentation

When to Use:

  • Creating or updating architecture documentation
  • Generating C4 diagrams (Context, Container, Component, Code)
  • Creating Mermaid sequence or flow diagrams
  • Documenting API endpoints and schemas
  • Maintaining ARCHITECTURE.md and SECURITY_ARCHITECTURE.md
  • Multi-language documentation strategies

Key Capabilities:

  • C4 architecture model implementation
  • Mermaid diagram generation
  • European Parliament MCP API documentation
  • ISMS policy documentation
  • Multi-language documentation patterns (14 languages)
  • README.md best practices
  • Architecture decision records (ADRs)

Example Use:

@documentation-architect create a C4 Container diagram showing the European Parliament MCP integration architecture

7. ✅ Quality Engineer (quality-engineer)

Expertise: Testing, validation, accessibility testing, performance benchmarking

When to Use:

  • Writing or improving tests
  • Validating HTML/CSS with automated tools
  • Conducting accessibility testing (WCAG 2.1 AA)
  • Performance benchmarking and optimization
  • Link integrity checking
  • Multi-language quality assurance

Key Capabilities:

  • Playwright visual regression testing
  • HTMLHint and CSSLint validation
  • WCAG 2.1 AA accessibility testing
  • Core Web Vitals measurement
  • Link integrity verification (linkinator)
  • Test coverage and mutation testing
  • Multi-language QA (14 languages)
  • Cross-browser testing strategies

Example Use:

@quality-engineer create Playwright tests to validate all 14 language versions for accessibility compliance

8. 🔍 Intelligence Operative (intelligence-operative)

Expertise: Political science, intelligence analysis, OSINT, behavioral analysis, EU Parliament transparency

When to Use:

  • Analyzing MEP voting patterns and political group dynamics
  • Conducting structured analytic techniques (ACH, SWOT, Devil's Advocacy)
  • Assessing democratic accountability and institutional performance
  • Evaluating legislative outcomes and coalition stability
  • Producing data-driven intelligence assessments from EP MCP data
  • Risk assessment for EU political developments

Key Capabilities:

  • Political science frameworks (comparative politics, public policy)
  • OSINT methodologies for EU Parliament open data
  • Intelligence analysis techniques (ACH, PESTLE, stakeholder analysis)
  • MEP scorecards and political group analysis
  • Electoral analysis and coalition dynamics
  • Behavioral analysis for MEP decision-making patterns
  • European Parliament MCP data integration
  • GDPR-compliant political data analysis

Example Use:

@intelligence-operative analyze voting cohesion trends across political groups using EP MCP voting records and produce a risk assessment

9. 💼 Business Development Specialist (business-development-specialist)

Expertise: Strategic planning, partnership development, revenue models, civic tech sustainability

When to Use:

  • Developing strategic growth plans for the platform
  • Identifying partnership opportunities with EU institutions, media, NGOs
  • Designing mission-aligned revenue models (grants, consulting, open core)
  • Analyzing market segments across 27 EU member states
  • Creating investor/grant proposals for civic tech funding
  • Evaluating business model viability and sustainability

Key Capabilities:

  • Business Model Canvas for civic tech platforms
  • Partnership development (EU institutions, media, academic, NGO)
  • Revenue strategy (EU Horizon grants, democracy funds, consulting)
  • Market segmentation across 27 EU member states
  • Go-to-market strategy for political transparency platforms
  • Financial modeling and sustainability planning
  • ISMS compliance for business decisions
  • GDPR-compliant business practices

Example Use:

@business-development-specialist create a Business Model Canvas for EU Parliament Monitor with focus on grant funding and institutional partnerships

10. 📢 Marketing Specialist (marketing-specialist)

Expertise: Digital marketing, content strategy, SEO, community building, multi-language outreach

When to Use:

  • Developing SEO strategy for multi-language news articles
  • Creating content marketing plans for EU Parliament coverage
  • Building community engagement across EU member states
  • Optimizing search visibility for 14 language versions
  • Designing social media campaigns for democratic engagement
  • Measuring marketing effectiveness and audience growth

Key Capabilities:

  • Multi-language SEO optimization (14 languages, hreflang)
  • Content strategy for EU Parliament journalism
  • Social media marketing for civic engagement
  • Community building across EU member states
  • Brand positioning for transparency platforms
  • Performance measurement and analytics
  • GDPR-compliant marketing (privacy-first, no tracking cookies)
  • Political neutrality in all messaging
  • Structured data (JSON-LD) and Schema.org optimization

Example Use:

@marketing-specialist develop a multi-language SEO strategy for EU Parliament news articles targeting all 14 language versions

🔧 GitHub Agentic Workflows (gh-aw) Agents

The following agents are sourced from github/gh-aw and provide GitHub Agentic Workflows capabilities:

11. 🤖 Agentic Workflows (agentic-workflows.agent)

Expertise: GitHub Agentic Workflows (gh-aw) — Create, debug, and upgrade AI-powered workflows

When to Use:

  • Creating new agentic workflows in markdown
  • Debugging failing workflows
  • Upgrading workflows to new gh-aw versions
  • Creating shared workflow components
  • Fixing Dependabot PRs for workflow dependencies
  • Analyzing test coverage

12. 🧹 CI Cleaner (ci-cleaner.agent)

Expertise: Repository CI state cleanup — formatting, linting, testing, and workflow recompilation

When to Use:

  • CI is failing and needs cleanup
  • Code formatting needs fixing
  • Linters report issues
  • Tests are broken
  • Workflows need recompilation

13. ✅ Contribution Checker (contribution-checker.agent)

Expertise: PR evaluation against CONTRIBUTING.md guidelines

When to Use:

  • Evaluating PR compliance with contribution guidelines
  • Assessing PR quality (on-topic, focused, tested, described)
  • Generating structured PR feedback

14. 🔌 Create Safe Output Type (create-safe-output-type.agent)

Expertise: Adding new safe output types to GitHub Agentic Workflows

When to Use:

  • Adding a new safe output type to the gh-aw system
  • Implementing JSONL validation pipelines
  • Creating handler factories for safe outputs

15. ⚙️ Custom Engine Implementation (custom-engine-implementation.agent)

Expertise: Implementing custom agentic engines in gh-aw

When to Use:

  • Building a new AI engine for gh-aw
  • Understanding the engine interface architecture
  • Integrating engines with MCP servers and firewalls

16. 📖 Developer Instructions (developer.instructions)

Expertise: Development guidelines and standards for GitHub Agentic Workflows

When to Use:

  • Understanding code organization patterns
  • Following validation architecture
  • Applying security best practices
  • Managing releases with changesets

17. 🔥 Grumpy Reviewer (grumpy-reviewer.agent)

Expertise: Thorough code review with a grumpy, sarcastic senior developer persona

When to Use:

  • Getting brutally honest code reviews
  • Finding code smells, performance issues, and security concerns
  • Identifying missing error handling and poor naming

18. 🎯 Interactive Agent Designer (interactive-agent-designer.agent)

Expertise: Interactive wizard for creating and optimizing agent prompts and workflow descriptions

When to Use:

  • Creating new agent prompts via guided wizard
  • Optimizing existing workflow descriptions
  • Designing workflow configurations (frontmatter)
  • Creating custom agent instructions

19. 📝 Technical Doc Writer (technical-doc-writer.agent)

Expertise: Technical documentation using GitHub Docs voice and Diátaxis structure

When to Use:

  • Writing developer-focused documentation
  • Creating getting-started guides, how-to guides, and references
  • Following GitHub Docs style and tone

20. 📋 W3C Specification Writer (w3c-specification-writer.agent)

Expertise: Formal W3C-style specifications with RFC 2119 keywords

When to Use:

  • Writing formal technical specifications
  • Creating protocol, API, or data format specifications
  • Applying RFC 2119 requirement levels (MUST, SHALL, SHOULD, MAY)
  • Defining conformance classes and compliance testing

21. 📥 News Generation (imported) (news-generation.agent)

Type: gh-aw imported agent — body is appended to every article-generating news-*.md workflow prompt via the imports: field. Not invokable directly via @news-generation.

Purpose: Contributes the canonical Required Reading order and 5-stage Stage Contract (Data → Analysis → Completeness Gate → Article → Single PR) to every importing workflow. Paired with .github/workflows/shared/mcp/news-mcp-servers.md (separate import that merges the mcp-servers: frontmatter block).

Importers: news-breaking, news-week-in-review, news-month-in-review, news-week-ahead, news-month-ahead, news-committee-reports, news-motions, news-propositions (the 8 unified article workflows). Explicitly not imported by news-translate (multi-call flush pattern, exempt from single-PR rule). The previous split-pair news-<type>-analysis.md

  • news-<type>-article.md layout and the manual news-article-generator.md helper were removed in the April-2026 aggregator-pipeline migration.

See .github/workflows/README.md for the full shared-import pattern.

🌍 European Parliament Context

All agents are configured with expertise in:

Data Sources

  • European-Parliament-MCP-Server (6 tools):
    • get_meps - MEP profiles and data
    • get_plenary_sessions - Session schedules and agendas
    • search_documents - Parliamentary document search
    • get_parliamentary_questions - Questions and answers
    • get_committee_info - Committee details and members
    • get_voting_records - Voting results and patterns

Multi-Language Support (14 Languages)

  • Nordic: English (en), Swedish (sv), Danish (da), Norwegian (no), Finnish (fi)
  • EU Core: German (de), French (fr), Spanish (es), Dutch (nl)
  • Middle East: Arabic (ar), Hebrew (he)
  • East Asia: Japanese (ja), Korean (ko), Chinese (zh)

Article Types

  • Week Ahead: Preview of upcoming plenary sessions and events
  • Committee Reports: Analysis of committee activities and decisions
  • Propositions: Government and parliamentary legislative proposals
  • Motions: Parliamentary motions and resolutions

🔒 ISMS Compliance Framework

All agents enforce compliance with the Hack23 ISMS. The three foundational SDLC policies every agent honours:

Foundational SDLC Policies

Policy Scope Enforced by
Information Security Policy CIA triad, risk management, ownership, least privilege, defense in depth All agents; primary in product-task-agent, documentation-architect
Secure Development Policy SSDLC phase gates, threat modelling, secure coding, testing, release devops-engineer, data-pipeline-specialist, quality-engineer, frontend-specialist
Open Source Policy License intake, SBOM, SLSA, contribution workflow, supply chain devops-engineer, data-pipeline-specialist, product-task-agent

Consolidated checklist + phase gates: sdlc-security-integration skill

ISO 27001:2022 Controls

  • A.5.10: Information use (European Parliament transparency)
  • A.8.3: Access restrictions (GitHub permissions, branch protection)
  • A.8.23: Web filtering (CSP headers, security policies)
  • A.8.24: Cryptography (TLS 1.3, HTTPS-only)
  • A.8.25: Secure SDLC (enforced via sdlc-security-integration skill)
  • A.8.28: Secure coding (HTML/CSS validation, input sanitization)

NIST CSF 2.0 Functions

  • Govern: ISMS policies, risk tolerance, ownership (product-task-agent)
  • Identify: Asset inventory (repositories, domains, data sources)
  • Protect: Access control (GitHub MFA, branch protection)
  • Detect: Monitoring (Dependabot, CodeQL, audit logs)
  • Respond: Incident procedures (rollback, hotfix)
  • Recover: Recovery planning (git history, backups)

CIS Controls v8.1

  • Control 1: Asset inventory
  • Control 4: Secure configuration (GitHub Pages, headers)
  • Control 6: Access control (branch protection, MFA)
  • Control 7: Continuous vulnerability management (Dependabot, CodeQL)
  • Control 8: Audit logging (GitHub audit logs)
  • Control 14: Security awareness (agent documentation)
  • Control 15: Service provider management (Third_Party_Management.md)
  • Control 16: Application security (validation, SAST)

EU Regulations

  • GDPR: Data protection, privacy by design
  • NIS2: Network and information security directive
  • EU CRA: Cyber Resilience Act compliance (SBOM, SLSA L3, vuln disclosure)

Reference: Hack23 ISMS-PUBLIC Repository

🤖 GitHub Copilot Insiders Features

All agents support modern Copilot coding agent features:

1. Basic Issue Assignment

// Assign Copilot to an issue
await github.issue_write({
  method: "update",
  owner: "Hack23",
  repo: "euparliamentmonitor",
  issue_number: 42,
  assignees: ["copilot-swe-agent[bot]"]
});

2. Feature Branch Assignment

// Work from a feature branch
await github.assign_copilot_to_issue({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  issue_number: 42,
  base_ref: "feature/mep-profiles"
});

3. Custom Instructions Assignment

// Provide additional context
await github.assign_copilot_to_issue({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  issue_number: 42,
  base_ref: "main",
  custom_instructions: `
    - Follow European Parliament MCP patterns
    - Support all 14 languages
    - Ensure WCAG 2.1 AA compliance
    - Reference ISMS policies
  `
});

4. Direct PR Creation

// Create PR with specific agent
await github.create_pull_request_with_copilot({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  title: "Add MEP voting visualization",
  body: "Implement interactive voting charts using European Parliament MCP data",
  base_ref: "main",
  custom_agent: "data-pipeline-specialist"
});

5. Stacked PRs

// Sequential PRs building on each other
const pr1 = await github.create_pull_request_with_copilot({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  title: "Step 1: Add MEP data layer",
  body: "Implement European Parliament MCP client",
  base_ref: "main"
});

const pr2 = await github.create_pull_request_with_copilot({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  title: "Step 2: Add MEP profiles",
  body: "Create multi-language MEP profile pages",
  base_ref: pr1.branch  // Stack on PR 1
});

6. Job Status Tracking

// Monitor Copilot progress
const status = await github.get_copilot_job_status({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  job_id: "abc123-def456"
});

🔗 Cross-Repository Access

All agents have access to the following Hack23 organization repositories:

Repository Purpose
European-Parliament-MCP-Server MCP server implementation, API schemas, tool documentation
riksdagsmonitor Similar static site patterns, news generation workflows
cia OSINT methodologies, intelligence analysis patterns
ISMS-PUBLIC Compliance policies (ISO 27001, GDPR, NIS2), security requirements
homepage Translation guides, multi-language best practices

Access Pattern Example:

// Reference riksdagsmonitor patterns
const riksdagPatterns = await github.get_file_contents({
  owner: "Hack23",
  repo: "riksdagsmonitor",
  path: "src/aggregator/article-generator.ts"
});

// Check ISMS policies
const secureDevPolicy = await github.get_file_contents({
  owner: "Hack23",
  repo: "ISMS-PUBLIC",
  path: "Secure_Development_Policy.md"
});

📊 Agent Selection Guide

Use this decision tree to select the right agent:

graph TD
    A[What type of task?] --> B{Category}
    
    B -->|Issue Creation| C[product-task-agent]
    B -->|Content Writing| D[news-journalist]
    B -->|UI/UX Work| E[frontend-specialist]
    B -->|Data Integration| F[data-pipeline-specialist]
    B -->|CI/CD| G[devops-engineer]
    B -->|Security| H[skills: security-by-design, threat-modeling, sdlc-security-integration]
    B -->|Documentation| I[documentation-architect]
    B -->|Testing/QA| J[quality-engineer]
    B -->|Intelligence Analysis| K2[intelligence-operative]
    B -->|Business Strategy| L2[business-development-specialist]
    B -->|Marketing/SEO| M2[marketing-specialist]
    
    C --> K[Creates issues with proper labels/priorities]
    D --> L[Writes The Economist-style articles]
    E --> M[Implements WCAG 2.1 AA UI]
    F --> N[Integrates European Parliament MCP]
    G --> O[Automates workflows]
    H --> P[Hardens security]
    I --> Q[Documents architecture]
    J --> R[Validates quality]
    K2 --> S[Produces intelligence assessments]
    L2 --> T[Develops growth strategy]
    M2 --> U[Optimizes reach and engagement]
Loading

Quick Reference Table

Task Type Primary Agent Secondary Agent(s)
Create product issues product-task-agent -
Write news articles news-journalist data-pipeline-specialist, intelligence-operative
Fix accessibility frontend-specialist quality-engineer
Add MEP data data-pipeline-specialist frontend-specialist
Setup CI/CD devops-engineer skills: security-by-design, sdlc-security-integration
Security audit skills: security-by-design, threat-modeling quality-engineer, grumpy-reviewer.agent
Architecture docs documentation-architect -
Run tests quality-engineer frontend-specialist
Political analysis intelligence-operative news-journalist, data-pipeline-specialist
Business planning business-development-specialist marketing-specialist
SEO optimization marketing-specialist news-journalist, frontend-specialist
Grant proposals business-development-specialist intelligence-operative
Audience growth marketing-specialist business-development-specialist

🚀 Usage Examples

Example 1: Create Feature Issue

# Use product-task-agent to analyze and create issues
@product-task-agent analyze the European Parliament MCP integration and create issues for any missing features or quality gaps

Expected Outcome: Multiple well-structured GitHub issues with:

  • Clear titles and descriptions
  • Proper labels (type:feature, priority:high, etc.)
  • Acceptance criteria
  • ISMS compliance mapping
  • Recommended agent assignments

Example 2: Generate News Article

# Use news-journalist to create content
@news-journalist create a week-ahead article for the upcoming plenary session in all 14 languages using European Parliament MCP data

Expected Outcome: Multi-language news articles with:

  • The Economist editorial style
  • European Parliament MCP data integration
  • SEO-optimized metadata
  • Proper HTML5 semantic structure
  • GDPR-compliant content

Example 3: Implement Accessibility Fix

# Use frontend-specialist for UI improvements
@frontend-specialist make the language switcher fully keyboard accessible and ensure it passes WCAG 2.1 AA Level A compliance

Expected Outcome: Accessible UI component with:

  • Keyboard navigation (Tab, Enter, Space, Arrow keys)
  • ARIA attributes for screen readers
  • Focus indicators
  • Playwright tests for keyboard interaction
  • WCAG 2.1 AA validation

Example 4: Setup Automated Testing

# Use devops-engineer for CI/CD
@devops-engineer create a GitHub Actions workflow to validate HTML/CSS on every pull request and run Playwright accessibility tests

Expected Outcome: GitHub Actions workflow with:

  • HTMLHint and CSSLint validation
  • Playwright browser testing
  • Accessibility auditing
  • Multi-language testing
  • Results reporting in PR comments

Example 5: Political Intelligence Assessment

# Use intelligence-operative for deep analysis
@intelligence-operative analyze political group cohesion trends using EP MCP voting records and produce a SWOT assessment for cross-group collaboration

Expected Outcome: Intelligence product with:

  • Structured analytic techniques (ACH, SWOT)
  • Data-driven analysis from EP MCP voting records
  • Political group cohesion metrics
  • Risk assessment and early warning indicators
  • GDPR-compliant political data analysis

Example 6: Business Strategy Planning

# Use business-development-specialist for growth planning
@business-development-specialist create a Business Model Canvas for EU Parliament Monitor with focus on EU Horizon grant opportunities

Expected Outcome: Strategic business document with:

  • Business Model Canvas framework
  • Revenue stream analysis (grants, consulting, partnerships)
  • Customer segmentation across 27 EU member states
  • Partnership strategy for EU institutions
  • Mission alignment with democratic transparency

Example 7: SEO and Marketing Strategy

# Use marketing-specialist for audience growth
@marketing-specialist develop a multi-language SEO strategy for EU Parliament news articles targeting all 14 language versions

Expected Outcome: Marketing strategy with:

  • Multi-language SEO keyword research
  • Hreflang implementation guidance
  • JSON-LD structured data optimization
  • Core Web Vitals targets
  • GDPR-compliant analytics approach
  • Political neutrality in all messaging

🏗️ Agent Development Patterns

Pattern 1: Coordinated Multi-Agent Workflow

# Step 1: Product planning
@product-task-agent create a feature plan for MEP voting record visualization

# Step 2: Intelligence analysis
@intelligence-operative analyze voting patterns and identify newsworthy trends from EP MCP data

# Step 3: Data integration
@data-pipeline-specialist implement the European Parliament MCP integration for voting records

# Step 4: Frontend implementation
@frontend-specialist create the responsive voting chart UI with WCAG 2.1 AA support

# Step 5: Content creation
@news-journalist write data-driven articles based on intelligence analysis findings

# Step 6: SEO optimization
@marketing-specialist optimize articles for search visibility across 14 languages

# Step 7: Quality assurance
@quality-engineer write Playwright tests and validate accessibility across all 14 languages

# Step 8: Documentation
@documentation-architect create C4 diagrams and API documentation for the voting visualization feature

Pattern 2: Stacked PRs for Complex Features

// PR 1: Data layer (data-pipeline-specialist)
const pr1 = await github.create_pull_request_with_copilot({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  title: "Step 1: Add voting record data fetching",
  body: "Implement European Parliament MCP client for voting data",
  base_ref: "main",
  custom_agent: "data-pipeline-specialist"
});

// PR 2: UI layer (frontend-specialist)
const pr2 = await github.create_pull_request_with_copilot({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  title: "Step 2: Add voting chart visualization",
  body: "Create responsive voting chart using data from PR #1",
  base_ref: pr1.branch,
  custom_agent: "frontend-specialist"
});

// PR 3: Testing (quality-engineer)
const pr3 = await github.create_pull_request_with_copilot({
  owner: "Hack23",
  repo: "euparliamentmonitor",
  title: "Step 3: Add comprehensive tests",
  body: "Implement Playwright tests and accessibility validation",
  base_ref: pr2.branch,
  custom_agent: "quality-engineer"
});

📚 Additional Resources

Hack23 Organization

European Parliament

GitHub Copilot

Standards & Compliance

🛠️ Development & Maintenance

Adding New Agents

When creating new agents, follow this structure:

---
name: agent-name
description: Brief description (max 200 chars)
# tools: omit this field — Copilot CLI treats omission as equivalent to all tools.
# Never set tools: ["*"] inside an mcp-servers entry (awmg treats "*" as a literal tool name).
---

# Agent Title

## 📋 Required Context Files
## Role Definition
## Core Expertise
## Standards and Guidelines
## GitHub MCP Insiders Experimental Features
## Capabilities
## Boundaries & Limitations
## Integration with Other Agents
## Skills to Leverage
## Cross-Repository Access
## Quality Standards
## Remember

Updating Agents

When updating agents:

  1. Maintain YAML frontmatter structure
  2. Ensure GitHub MCP Insiders features are documented
  3. Keep ISMS compliance mappings current
  4. Update cross-repository references as needed
  5. Test agent behavior in GitHub Copilot
  6. Document changes in commit messages

🤝 Contributing

When contributing to agent development:

  1. Follow Standards: Use existing agents as templates
  2. Test Thoroughly: Validate YAML syntax and agent behavior
  3. Document Fully: Include examples and use cases
  4. ISMS Compliance: Map to ISO 27001, NIST CSF, CIS Controls
  5. European Parliament Focus: Ensure EU-specific context
  6. Multi-Language: Support all 14 languages
  7. Security First: Follow Hack23 Secure Development Policy

📞 Support

For questions or issues:

Last Updated: 2026-02-16
Version: 1.0
Maintained by: Hack23 AB