Back to Guides

A short AI-narrated film with English captions. Pressing play loads the YouTube player. Watch on YouTube

Guide Origin: Official | ArcKit Version: 4.4.0

/arckit:gov-landscape maps what UK government has built in a domain — organisations active in the space, common technology patterns, adopted standards, maturity levels, and collaboration opportunities — giving architects a broad situational picture before making design decisions.

Agent Architecture: This command runs as an autonomous agent via the Task tool. The agent issues 8-12 govreposcrape queries across the domain, uses WebFetch to inspect representative repositories at depth, and synthesises findings into a structured landscape report, keeping the main conversation clean. The slash command is a thin wrapper that delegates to the agent.

Prerequisites

govreposcrape MCP Server (AUTO-INSTALLED)

This command uses the govreposcrape MCP server which is automatically included with the ArcKit plugin. No manual setup required.

Endpoint: https://govreposcrape-api-1060386346356.us-central1.run.app/mcp

Project Prerequisites

  • Requirements document (ARC-*-REQ-*.md) - Recommended (used to anchor domain focus)
  • Architecture principles (ARC-000-PRIN-*.md) - Recommended
  • Stakeholder analysis (ARC-*-STKE-*.md) - Optional

Scenario Matrix

Scenario Prompt seed Focus
Domain mapping "Map the government landscape for <domain>" Breadth of activity across departments
Technology survey "Survey technology used in government <domain>" Language, framework, and platform spread
Standards adoption "Which government teams have adopted <standard>?" Compliance patterns and gaps
Collaboration discovery "Find government teams working on <problem>" Potential reuse partners or communities of practice
Gap analysis "What is missing from the government <domain> landscape?" Under-served capabilities and build opportunities

Add a time qualifier ("in the last 2 years") or a department name to scope the landscape further.

Command

/arckit:gov-landscape <domain>

Outputs: projects/<id>/research/ARC-<id>-GLND-v1.0.md

Auto-versioning: Re-running this command when a document already exists automatically increments the version.

How It Works

The agent uses a broad-to-specific search strategy across the govreposcrape semantic search API:

Step Action
1 Generate 8-12 domain-spanning queries covering different facets
2 Run all queries, collecting a wide pool of repositories
3 Cluster results by organisation and technology theme
4 Fetch GitHub repo details via WebFetch for representative samples in each cluster
5 Score each organisation's activity on domain maturity (depth, recency, documentation)
6 Synthesise findings into a landscape map with technology breakdown and gap analysis

Maturity Scoring

Each active organisation in the domain is rated on three dimensions:

Dimension High Low
Depth Multiple repos, clear investment Single exploratory repo
Recency Active commits within 12 months No activity in 2+ years
Documentation Architecture docs, guides, READMEs Minimal or no documentation

Output Highlights

  • Landscape map showing organisations active in the domain and their relative maturity
  • Technology stack analysis with language and framework distribution across repos
  • Standards adoption summary — which standards are widely, partially, or not yet adopted
  • Maturity scores per organisation for depth, recency, and documentation
  • Collaboration opportunities identifying teams solving similar problems
  • Gap analysis for capabilities or standards with little or no existing government code

Follow-on Actions

  • Run /arckit:gov-reuse to perform a scored reuse assessment for specific capabilities identified in the landscape
  • Run /arckit:framework to transform landscape findings into architectural principles and patterns
  • Run /arckit:wardley to map domain components onto a Wardley Map for evolution analysis
Feature /arckit:gov-landscape /arckit:gov-reuse /arckit:gov-code-search
Purpose Map an entire domain Assess candidates for a project Discover repos matching a query
Scope Domain-wide (8-12 queries) Capability-specific (3-5 queries) Query-specific (3-5 queries)
Output Landscape map + maturity scores Scored reuse assessment per capability Search results + repo profiles
Project required No (recommended) Yes (requirements MANDATORY) No
When to use Early-stage situational awareness Before building a capability Exploratory or ad-hoc search

Workflow: Use /arckit:gov-landscape to understand the domain, /arckit:gov-code-search for targeted queries, and /arckit:gov-reuse for project-specific reuse decisions.

Resources