Architecture Workflow logo

Architecture Workflow

CommunityPopular
tractorjuice
architecture-workflow

Recommends which /arckit:* commands to run, and in what order, for a new or in-flight architecture project, from a short triage of sector, project type, current stage and timeline. Backs /arckit:start. Not needed when the user has already named the command they want, or is asking about the content of an artefact rather than the sequence.

Overview

Publishertractorjuice
Repositoryarc-kit
Skill namearchitecture-workflow
Stars
2.2K
Forks
280
Bundled files
5
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 5 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by tractorjuice on GitHub. Read the source before you install it.

Installation

Install the Architecture Workflow AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/tractorjuice/arc-kit.git /tmp/arc-kit
mkdir -p .claude/skills
cp -r /tmp/arc-kit/plugins/arckit-claude/skills/architecture-workflow .claude/skills/architecture-workflow
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Architecture Workflow in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Architecture Workflow on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Architecture Workflow is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Architecture Workflow

Guides users through project onboarding using adaptive-depth questions and recommends a tailored command sequence.

Anti-Patterns

"I already know what I need"

Even experienced architects benefit from the triage. It catches blind spots — missing compliance requirements, forgotten dependencies, stakeholder gaps. The triage is fast (3-4 questions). Skip it and you risk generating artifacts in the wrong order or missing mandatory prerequisites.

"Just run everything"

A 30-command sequence helps nobody. The skill's job is to recommend the right commands for this project, in the right order. Every project is different — a compliance review needs 6 commands, not 30.

Process

Follow these steps in order. Ask questions one at a time using AskUserQuestion. Prefer multiple-choice options.

Step 1: Detect Project State

Automatically check the project context (no questions needed):

  • Check if projects/ directory exists and count projects
  • Check for principles document (ARC-000-PRIN-*)
  • Count existing artifacts per project
  • Use ArcKit Project Context from the SessionStart hook if available

Based on findings, determine:

  • New project: No projects/ directory or empty — recommend starting from scratch
  • Early stage: Projects exist but few artifacts (0-24% complete) — recommend next foundation steps
  • Mid stage: Has requirements, some design artifacts (25-74%) — recommend design and procurement steps
  • Late stage: Has most artifacts (75-100%) — recommend quality, compliance, and reporting steps

Display a brief status summary before asking questions:

text
Project State: [1 project found, 4 artifacts, ~20% complete]

Or:

text
Project State: No project structure found. Starting fresh.

Step 2: Triage Questions

Ask these questions one at a time. Each uses AskUserQuestion with multiple-choice options.

Question 1 — Sector:

  • UK Government (civilian departments)
  • Defence (MOD, defence contractors)
  • Public sector (non-UK)
  • Private sector

Question 2 — Project Type:

  • New system build
  • System migration or modernization
  • Procurement / vendor selection
  • Data platform or analytics
  • AI/ML system
  • Strategy or governance review only

Question 3 — Current Stage:

  • Just starting (no artifacts yet)
  • Have stakeholders and/or requirements
  • Have design artifacts (data model, research, diagrams)
  • Need compliance review of existing work

Question 4 — Primary Goal:

  • Full governance lifecycle (end-to-end)
  • Specific deliverable (e.g., just requirements, just SOBC)
  • Compliance check (assess existing work)
  • Quick prototype documentation (minimum viable)

Step 3: Deep Questions (Complex Projects Only)

Only ask these if the project triggers complexity:

  • Sector is UK Government or Defence
  • Project type is AI/ML
  • Primary goal is full governance lifecycle

Ask one at a time:

Q5 — Compliance Frameworks (multiple select):

  • GDS Service Standard
  • Technology Code of Practice (TCoP)
  • NCSC Cyber Assessment Framework
  • AI Playbook
  • JSP 440 / MOD Secure by Design
  • JSP 936 / MOD AI Assurance
  • None / not sure

Q6 — Procurement (if applicable):

  • G-Cloud (Digital Marketplace)
  • Digital Outcomes and Specialists (DOS)
  • Open tender / framework agreement
  • No procurement needed

Q7 — Strategic Analysis:

  • Yes, need Wardley Maps and strategic positioning
  • Yes, need platform design (multi-sided platform)
  • No, straightforward technology choices

Q8 — Timeline Pressure:

  • Weeks (urgent, minimum viable only)
  • Months (standard delivery)
  • Quarters (major programme, full governance)

Step 4: Present Tailored Plan

Based on the answers, select the appropriate path and present the plan.

Decision Logic

Base path selection:

Sector AnswerBase Path
Private sector or Public sector (non-UK)standard-path.md
UK Governmentuk-gov-path.md
Defencedefence-path.md

Modifiers (applied on top of base path):

ConditionModifier
Project type = AI/MLApply ai-ml-path.md
Project type = Data platformApply data-path.md
Both AI/ML and DataApply both modifiers

Scope adjustments:

GoalAdjustment
Full governance lifecycleShow full path from base + modifiers
Specific deliverableShow only the relevant phase
Compliance checkShow only compliance phase from base path
Quick prototype documentationShow minimum viable path from base path
StageAdjustment
Just startingShow full path (or scoped path)
Have stakeholders/requirementsSkip Phases 1-2, start from Phase 3
Have design artifactsSkip to Phase 4 (Procurement) or Phase 5 (Design Reviews)
Need compliance reviewSkip to compliance phase
TimelineAdjustment
WeeksShow minimum viable path only
MonthsShow full path, note optional commands
QuartersShow full path with all optional additions
Plan Output Format

Present the plan as a numbered list grouped by phase:

text
Recommended Command Sequence
=============================

Based on: [UK Government] + [AI/ML] project, starting fresh, full lifecycle

Phase 1: Foundation
  1. /arckit:principles — Governance foundation (GDS + TCoP aligned)
  2. /arckit:stakeholders — Map DDaT roles, SROs, policy owners
  3. /arckit:risk — HMG Orange Book risk methodology

Phase 2: Business Justification
  4. /arckit:sobc — HM Treasury Green Book 5-case model
  5. /arckit:requirements — Central artifact for all downstream work

Phase 3: Design & Analysis
  6. /arckit:datascout — Discover UK Gov open data sources
  7. /arckit:data-model — Data architecture with GDPR considerations
  ...

[Total: N commands across M phases]
[Estimated duration: X-Y months]

Run commands in order. Each command will guide you through its process.

After presenting the plan, ask if they want to adjust anything or if they're ready to begin.

Key Principles

  • One question at a time — do not overwhelm with multiple questions per message
  • Multiple choice preferred — easier to answer than open-ended
  • Adaptive depth — simple projects get 4 questions, complex get 8
  • Scaled output — minimum viable = 5 commands, full lifecycle = 25-30
  • No commands executed — only present the plan, user drives execution
  • Reference existing artifacts — skip phases where artifacts already exist

ArcKit Integration

This skill is invoked by the /arckit:start command, which delegates project onboarding to this skill. Users can also trigger it by asking about getting started, command sequences, or workflow recommendations.

For the detailed command dependency matrix, see DEPENDENCY-MATRIX.md in the user's project root (installed by arckit init). For visual workflow diagrams, see WORKFLOW-DIAGRAMS.md in the user's project root.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Architecture Workflow AI skill do?

Recommends which /arckit:* commands to run, and in what order, for a new or in-flight architecture project, from a short triage of sector, project type, current stage and timeline. Backs /arckit:start. Not needed when the user has already named the command they want, or is asking about the content of an artefact rather than the sequence.

Why use Architecture Workflow on TypingMind?

Because you install it once and use it with any model. Architecture Workflow is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Architecture Workflow in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/tractorjuice/arc-kit/tree/main/plugins/arckit-claude/skills/architecture-workflow. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Architecture Workflow?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Architecture Workflow?

As many as you like. As long as a model supports skills, you can use Architecture Workflow with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Architecture Workflow AI skill free?

It is published on GitHub by tractorjuice. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇