Skip to main content
Defra Logo

GitHub Copilot Setup for Defra

Practical setup guidance for configuring GitHub Copilot in Defra projects — agents, instructions, skills, prompts, and cross-tool configuration.

LAP Implementations Agents - Examples for GitHub Copilot

Legacy Application Modernisation (LAP) agents are specialised AI personas you can invoke in GitHub Copilot Chat to support the safe, incremental modernisation of legacy applications. Each agent has a defined role, expertise, and workflow that shapes how Copilot responds.

How agents work

When you select an agent in Copilot Chat (using @agent-name), Copilot adopts that persona for the conversation. The agent’s instructions override the default behaviour, giving you focused, role-specific assistance.

Agent files live in .github/agents/ and use the .agent.md extension.

How to use these examples

  1. Copy the .agent.md file into your repository’s .github/agents/ directory
  2. Edit any project-specific details (service name, tech stack, team conventions)
  3. Open Copilot Chat and select the agent from the agent picker

Keep agents current. Defra standards evolve. Revisit agent files quarterly or when the Defra software development standards (opens in new tab) repository is updated.

Available examples

Agent Purpose
LAP Innovation Application Developer Extracts workflows, behaviours, domain concepts, and business rules from legacy .NET and VB source code.
LAP Innovation Business Analyst Extracts ubiquitous language, bounded contexts, subdomains, and context maps from legacy application evidence.
LAP Innovation Cloud Tier Architect Applies Defra CCoE cloud service-tier requirements to the product requirements and architecture.
LAP Innovation Completeness Auditor Reconciles the build against the feature traceability manifest so functionality is not lost.
LAP Innovation Database Analyst Extracts schema, stored procedure logic, triggers, constraints, and persistence patterns from legacy SQL Server code.
LAP Innovation Digital Content Curator Converts screenshots and interview transcripts into curated material for downstream analysis.
LAP Innovation Digital Content Processor Processes one raw file using a specified skill for the digital content curator.
LAP Innovation Feature Writer Writes one standards-compliant feature specification from the relevant product requirements content.
LAP Innovation Interaction Analyst Combines HTML mockups and curated transcripts into interaction analysis and user workflows.
LAP Innovation LAP Orchestrator Runs the end-to-end LAP modernisation pipeline from content curation through feature traceability.
LAP Innovation Modernisation Example Writer Writes a standards-compliant modernisation example from the project evidence and codebase.
LAP Innovation PRD to Features Decomposes a product requirements document into deliverable, traceable feature specifications.
LAP Innovation Product Manager Synthesises analysis outputs into a comprehensive product requirements document.
LAP Innovation Documentation Produces factual, evidence-based system documentation (HLD, LLD, ADRs, Runbook) for the current codebase. No refactoring.
LAP Innovation Implementation Implements one approved migration slice per pull request, aligned to the intelligent migration programme, with tests, documentation, and rollback awareness.
LAP Innovation Intelligent Migration Establish a repeatable, low-risk migration operating model that increases delivery success probability using AI-augmented teams.
LAP Inovation Modernise to Modular Monolith Design a realistic modernisation path that can be executed safely and incrementally.
LAP Innovation Testing Create a safety net that enables confident, incremental modernisation.

Writing your own agents

Follow this structure:

---
description: One-line summary of what this agent does
tools: [edit, execute, read, search, web, todos, thinking]
---

# Agent Name

## Role
What this agent is and its expertise.

## Workflow
1. Step one
2. Step two
3. Step three

## Rules
- What the agent must always do
- What the agent must never do

## References
- Links to standards or instruction files this agent follows

Tips:

  • Keep the description under 100 characters — it appears in the agent picker
  • Prefer tool sets (edit, execute, read, search, web, agent) over individual tool names — each set bundles all the related built-in tools so the agent has full Agent-mode capability without a long brittle list
  • Add todos for any multi-step agent — it gives the agent (and you) a visible progress checklist
  • For read-only review agents, omit edit and execute to enforce the principle of least privilege
  • VS Code silently ignores unknown tool names (opens in new tab), so MCP and extension tools can be listed safely
  • Reference instruction files rather than restating rules
  • Define clear boundaries — what the agent does and does not do

Back to Getting Started

Support links

All content is available under the Open Government Licence v3.0, except where otherwise stated