archipilot – Enterprise Architecture Copilot
A VS Code Chat Participant extension that turns your Obsidian architecture vaults into an interactive, TOGAF-aligned Enterprise Architecture assistant — powered by GitHub Copilot. Ask architecture questions, resolve decisions, update vault documents — all from Copilot Chat. How It Works
Typical workflow:
Features
Commands
Palette Commands
Command Cheat SheetA quick reference for every
Flags: Append Usage ExamplesBelow are practical examples for every command. Just type these in Copilot Chat — the agent reads your vault and responds accordingly.
|
| Color | Meaning | Description |
|---|---|---|
| 🔴 Red | Remove | Exists in As-Is but not in Target — to be retired |
| 🟢 Green | Add | New in Target — to be implemented |
| 🔵 Blue | Keep | Unchanged — carried over from As-Is to Target |
Output files (saved to exports/drawio/ in your vault):
- Combined multi-tab
.drawiofile (all 3 views) - Individual
.drawiofiles for As-Is, Target, and Migration
Open in VS Code (with the "Draw.io Integration" extension), diagrams.net, or Confluence.
@architect /todo — TOGAF Action Items
Scan the vault and extract a prioritised list of next actions from a TOGAF perspective. The extractor scans:
- Open Decisions — from X1_ADR_Decision_Log.md
- Open Risks — from X2_Risk_Issue_Register.md
- Open Questions — from X3_Open_Questions.md
- Pending Work Packages & Milestones — from F1_Architecture_Roadmap.md
- Open Requirements — from R1_Architecture_Requirements.md (FR + NFR)
- Incomplete Compliance Checks — from G1_Compliance_Assessment.md
- Pending Change Requests — from H1_Change_Request_Log.md
- Draft / Review Documents — from YAML front matter status
- Unassigned Ownership — items with owner = TBD
@architect /todo
@architect /todo What should we focus on this sprint?
@architect /todo Which open decisions are blocking work packages?
@architect /todo --no-analysis (structured list only, skip AI prioritisation)
The AI analysis adds priority validation, sequencing advice, blocker identification, quick wins, and a suggested "next sprint" action list.
@architect — Free-form Questions
Ask anything about your architecture. The agent uses the full vault as context.
@architect What is the current state of our AI vendor strategy?
@architect Summarize all open decisions and their impact
@architect Which components depend on the API Gateway?
@architect What are the main risks if we choose Option C for AD-01?
@architect Explain the data flow between the Conversational Interface and the Rule Engine
@architect /analyze — Impact Analysis (Read-Only)
Assess the impact of a proposed change without modifying any file.
@architect /analyze What happens if we drop the API Gateway and use direct service-to-service calls?
@architect /analyze Impact of moving from single-vendor to multi-vendor LLM strategy
@architect /analyze What would change if we make this -only instead of enterprise-wide?
@architect /analyze Assess the effect of adding a real-time event bus to the integration layer
@architect /analyze How does removing DEP-03 (Rule Engine) affect our scenarios?
@architect /decide — Two-Step Decision Support
Decision support follows a two-step flow: first an analysis-only response with structured pros/cons, then — after your confirmation — the decision is recorded in the vault.
Step 1 — Analysis (no vault changes):
@architect /decide Help me resolve AD-01 — should we go with Feature, Enterprise Service, or AI Platform?
@architect /decide We need to pick between Kafka and RabbitMQ for the event bus. What does the architecture suggest?
@architect /decide Should we build or buy the Context Aggregation Engine?
Step 2 — Record (after reviewing the analysis, confirm to write):
Yes, let's go with Option B
Record this decision — Option A is the winner
The analysis step explicitly avoids generating any vault commands. Only after you confirm does archipilot emit an ADD_DECISION command with diff preview.
@architect /update — Modify Vault Documents
Tell the agent what to change in natural language. It generates structured commands and applies them to your files.
Add entries:
@architect /update Add a new dependency DEP-07: "Monitoring Platform" from Infrastructure, impact "Observability for all AI services"
@architect /update Add risk R-06: "Regulatory change" — probability Low, impact High, mitigation "Monitor EU AI Act updates quarterly"
@architect /update Add a new open question Q-25: "Do we need a dedicated ML platform team?" in category Governance, phase P
@architect /update Add decision AD-04: "Event-Driven vs Request-Response" with options Kafka, REST, Hybrid
@architect /update Add a new functional requirement FR-06: "AI provides multi-language support" priority Should, source Scenario 1, phase 3
Update entries:
@architect /update Change DEP-02 impact to "All external and internal routing, rate limiting, observability"
@architect /update Update risk R-04 probability from Medium to High — we got new latency test results
@architect /update Mark AD-01 as decided: Option B (Enterprise Service). Add rationale: "Aligns with platform strategy and reuse goals"
@architect /update Change the status of FR-01 from Open to In Progress
@architect /update Set the owner of Q-07 to "Security Team" with target date 2026-03-15
@architect /update Bump A1_Architecture_Vision to version 1.1.0 and status to "review"
Remove entries:
@architect /update Remove DEP-06 (MCP Strategy) from the dependencies — it's been moved to a separate initiative
@architect /update Remove assumption ASM-03, it has been validated
@architect /update Remove risk R-05 (Organizational resistance) — change management program approved
@architect /update Remove open question Q-12, answered: latency target is 500ms p95
Add new sections or content:
@architect /update Add a "Security Considerations" section to C1_Application_Architecture with subsections for AuthN, AuthZ, and Data Protection
@architect /update Add a rollback procedure for Phase 2 in F2_Migration_Plan
@architect /update Append a new building block "Prompt Safety Filter" to E1_Solutions_Building_Blocks — ABB: Content Safety, SBB: Azure AI Content Safety, decision: Buy
Create new files:
@architect /update Create a new file C4_API_Catalog.md with a catalog of all APIs identified in the architecture
@architect /update Create a new stakeholder interview notes file with today's meeting outcomes
@architect /status — Vault Overview
Get a dashboard of your architecture's health: document maturity, open decisions, risks, and questions.
@architect /status
@architect /status How many decisions are still open?
@architect /status Which documents are still in draft?
@architect /status Show me a summary of all risks by severity
@architect /status What percentage of requirements have been addressed?
@architect /switch — Change Project
Switch the active vault when working with multiple architecture projects.
@architect /switch
@architect /switch Change to the Payment-Platform vault
@architect /c4 — C4 Model Diagrams
Generate C4 architecture diagrams in Mermaid syntax from your vault content.
@architect /c4 Generate a System Context diagram for the AI Assistant
@architect /c4 Create a Container diagram showing all components and their interactions
@architect /c4 Component diagram for the Context Aggregation Engine
@architect /c4 Show the deployment view across cloud regions
@architect /sizing — Capacity & Cost Analysis
Analyze sizing, capacity planning, and cost estimation based on vault content.
@architect /sizing Estimate the compute requirements for 500 concurrent call center agents
@architect /sizing What would the monthly LLM API cost be at 10,000 interactions/day?
@architect /sizing Compare cost of self-hosted vs cloud LLM for our expected volume
@architect /sizing Update X4_Sizing_Catalogue with estimates for Phase 2 pilot
@architect /timeline — Gantt Timeline
Generate Mermaid Gantt charts from your roadmap and migration plan.
@architect /timeline Generate a Gantt chart from the roadmap in F1_Architecture_Roadmap
@architect /timeline Show the critical path for Phase 2 delivery
@architect /timeline Create a timeline for all 5 phases with dependencies
@architect /timeline What does the timeline look like if Phase 1 slips by 4 weeks?
@architect /new — Scaffold a New Vault
Create a fresh TOGAF-aligned vault with all 28 template files.
@architect /new Create a new vault called Payment-Platform-Architecture
@architect /new Scaffold a project for the Customer-360 initiative
@architect /scan — Generate Vault from Source Code
Scan the workspace source code and automatically populate a TOGAF-aligned vault. Detects tech stack, data models, services, infrastructure, API specs, and README content — then maps each signal to the correct TOGAF artifact (D2 standards catalog, C1 application components, C2 data entities, etc.).
Generate a new vault from scratch:
@architect /scan
- Scans workspace for package files, models, services, infrastructure, API specs
- Detects project name from
package.json,pyproject.toml, etc. - Creates a new TOGAF vault from the full 27-file template
- Calls the LLM to populate sections with real extracted information
- Shows a review step — extraction summary table with per-command validation, unified diffs, and error/warning indicators
- Commands with vault validation errors are automatically filtered out
- Apply valid commands with one click
Enrich an existing vault (append mode):
@architect /scan --append
Use this to add a second repo or microservice to an already-populated vault. Existing approved content is never overwritten — only new information is added.
What gets extracted:
| Source Signal | TOGAF Artifact Populated |
|---|---|
package.json, requirements.txt, pom.xml |
D2_Technology_Standards_Catalog.md — tech stack with versions |
openapi.yaml, swagger.json |
B2_Business_Capability_Catalog.md, C1_Application_Architecture.md |
models/, *.prisma, migrations/ |
C2_Data_Architecture.md — entities and data flows |
services/, controllers/, routes/ |
C1_Application_Architecture.md — components and interactions |
Dockerfile, docker-compose, *.tf |
D1_Technology_Architecture.md — deployment topology |
README.md, ARCHITECTURE.md |
A1_Architecture_Vision.md — executive summary and scope |
.env.example, config/*.yaml |
E2_Integration_Strategy.md — integration points |
| Technology choices across files | X1_ADR_Decision_Log.md — auto-generated decisions |
💡
/scanis the fastest way to bootstrap a TOGAF vault for an existing codebase. Run it once, then use/auditto find gaps and/updateto refine.
Dry-Run Mode
Append --dry-run or --preview to any /update command to see what would change without writing anything to disk:
@architect /update Add risk R-06: "Regulatory change" --dry-run
@architect /update Mark AD-01 as decided: Option B --preview
The preview now includes unified diffs (context-style --- a/file / +++ b/file) and vault-aware validation — the system checks file existence, section heading matches (with fuzzy suggestions for close matches), and duplicate detection before any write.
Supported Architecture Commands
When updating documents, the agent uses structured commands internally:
ADD_DECISION– Add decisions to the ADR Decision LogUPDATE_SECTION– Replace a section in any fileAPPEND_TO_FILE– Append content to a fileADD_OPEN_QUESTION– Add questions to the trackerCREATE_FILE– Create new vault documentsUPDATE_YAML_METADATA– Modify front-matter fields
Sidebar Features
The archipilot sidebar (click the AP icon in the Activity Bar) provides three views:
Vault Explorer
Browses vault files grouped by TOGAF ADM phase. Each file item has three inline action icons:
| Icon | Action |
|---|---|
| 📄 | Open the file in the editor |
| 💬 | Ask Copilot to analyze the file (/analyze) |
| 👁️ | Open the file in Split Preview (editor + rendered Markdown side-by-side) |
Search / Filter: Click the 🔍 icon in the Vault Explorer title bar (or run archipilot: Search / Filter Vault Files from the Command Palette) to filter files by name or content. Leave the input empty to clear the filter.
Quick Actions
Clickable shortcuts for the most common commands — no need to type in chat.
Customization: Hide actions you don't use by adding their label or /command shortcut to the archipilot.hiddenQuickActions setting. For example:
"archipilot.hiddenQuickActions": ["Sizing Catalogue", "/sizing", "Export to Draw.io"]
Architecture Health
A live linter that scans the vault and flags issues without running any AI:
| Check | What it flags |
|---|---|
| Orphaned Entities | Files with no incoming or outgoing [[WikiLinks]] |
| Risks without Owners | Rows in X2_Risk_Issue_Register.md where Owner is TBD or empty |
| Decisions Pending | Rows in X1_ADR_Decision_Log.md with status Proposed |
| Open TODOs | Inline TODO: markers across all vault files |
Click any item to jump directly to the file. Use the refresh button (↺) to re-scan after edits.
Getting Started
Prerequisites
- VS Code 1.93+
- GitHub Copilot extension (active subscription)
Installation
# Clone and build
cd archipilot
npm install
npm run compile
Running
- Press
F5to launch the Extension Development Host - Open a workspace containing your Obsidian vault folder(s)
- Open Copilot Chat (
Ctrl+Shift+I/Cmd+Shift+I) - Type
@architect /statusto start
Switching Projects
- Type
@architect /switchin chat, or - Click the status bar item "archipilot: …", or
- Run command palette → "archipilot: Switch Architecture Vault"
Export Output Structure
All diagram and model exports are saved into dedicated subfolders under exports/ in your vault:
My-Architecture-Vault/
├── exports/
│ ├── archimate/ ← ArchiMate 3.2 Open Exchange XML files
│ │ └── MyVault_ArchiMate_2026-02-17T10-30-00.xml
│ └── drawio/ ← Draw.io diagram files
│ ├── MyVault_DrawIO_2026-02-17T10-30-00.drawio (combined)
│ ├── MyVault_AsIs_2026-02-17T10-30-00.drawio
│ ├── MyVault_Target_2026-02-17T10-30-00.drawio
│ └── MyVault_Migration_2026-02-17T10-30-00.drawio
├── P1_Architecture_Principles.md
├── A1_Architecture_Vision.md
└── ...
Folders are created automatically on first export. C4 diagrams (/c4) output Mermaid syntax directly in chat.
Vault Structure — TOGAF ADM File Guide
The extension expects Obsidian vaults with markdown files following the TOGAF Architecture Development Method (ADM). Each file uses a prefix that maps to an ADM phase, making it instantly clear where every artifact belongs.
Any folder with 3+ .md files is auto-detected as a potential vault. Use @architect /new to scaffold a fresh vault with all files pre-created.
File Prefixes at a Glance
| Prefix | TOGAF Phase | Purpose |
|---|---|---|
00_ |
Index | Master index linking to all vault artifacts |
P* |
Preliminary | Principles and governance that apply before any cycle |
A* |
Phase A — Architecture Vision | Scope, stakeholders, value proposition |
B* |
Phase B — Business Architecture | Capabilities, scenarios, business processes |
C* |
Phase C — Information Systems | Application & data architecture |
D* |
Phase D — Technology Architecture | Infrastructure, platforms, standards |
E* |
Phase E — Opportunities & Solutions | Building blocks, integration, build-vs-buy |
F* |
Phase F — Migration Planning | Roadmap, work packages, transition states |
G* |
Phase G — Implementation Governance | Compliance, gate reviews, deviation tracking |
H* |
Phase H — Architecture Change Mgmt | Change requests against the approved architecture |
R* |
Requirements Management | Cross-phase requirements, constraints, assumptions |
X* |
Cross-phase | Decisions, risks, open questions, sizing — used throughout |
Complete File Reference
My-Architecture-Vault/
│
│── 00_Architecture_Repository.md ← Master index & navigation hub
│
│── P1_Architecture_Principles.md ← Guiding principles (rationale + implications)
│── P2_Governance_Framework.md ← Roles, review cadence, escalation paths
│
│── A1_Architecture_Vision.md ← Scope, drivers, constraints, high-level vision
│── A2_Stakeholder_Map.md ← Stakeholders, concerns, RACI, comms plan
│── A3_Value_Chain.md ← Value streams and business flow diagrams
│
│── B1_Business_Architecture.md ← Baseline → target business architecture & gap
│── B2_Business_Capability_Catalog.md ← Capability inventory with maturity levels
│── B3_Business_Scenarios.md ← Detailed scenarios driving architecture
│
│── C1_Application_Architecture.md ← Application landscape, interactions, gaps
│── C2_Data_Architecture.md ← Data entities, flows, governance policies
│── C3_Application_Portfolio_Catalog.md ← Full portfolio with lifecycle status
│
│── D1_Technology_Architecture.md ← Infrastructure, deployment topology
│── D2_Technology_Standards_Catalog.md ← Approved technologies & standards
│
│── E1_Solutions_Building_Blocks.md ← ABB → SBB mapping, build-vs-buy analysis
│── E2_Integration_Strategy.md ← Integration patterns, APIs, external deps
│
│── F1_Architecture_Roadmap.md ← Phased delivery plan with milestones
│── F2_Migration_Plan.md ← Transition states, rollback, data migration
│
│── G1_Compliance_Assessment.md ← Compliance checklist, gate reviews, deviations
│── G2_Architecture_Contracts.md ← Formal agreements between arch & impl teams
│
│── H1_Change_Request_Log.md ← Change requests against approved architecture
│
│── R1_Architecture_Requirements.md ← Functional, NFR, constraints, assumptions
│
│── X1_ADR_Decision_Log.md ← ADR-style architecture decisions
│── X2_Risk_Issue_Register.md ← Risks (probability × impact) + issues
│── X3_Open_Questions.md ← Questions tracked by phase & owner
│── X4_Sizing_Catalogue.md ← Component sizing, capacity, cost estimates
│── X5_Traceability_Matrix.md ← End-to-end traceability (driver → req → component → decision)
│── X6_Technical_Debt_Log.md ← Technical debt register with resolution tracking
How to Use Each File
00_Architecture_Repository.md — Start Here
The master index for the vault. Contains wiki-links ([[filename]]) to every artifact. Open this first to navigate the architecture. Update it whenever you add new files.
Preliminary (P*) — Set the Rules
- P1 Architecture Principles — Define the non-negotiable principles that guide all decisions. Each principle has a rationale (why) and implications (what it means in practice). Review these before making any AD decision.
- P2 Governance Framework — Who reviews what, how often, and how escalations work. Establish this early — it prevents circular debates later.
Phase A (A*) — Define the Vision
- A1 Architecture Vision — The "elevator pitch" for the architecture. Contains scope, business drivers, constraints, and the high-level target state. This is the first document stakeholders see.
- A2 Stakeholder Map — Everyone who has a say or is affected. Includes a RACI matrix and communication plan. Use it to ensure no stakeholder is surprised.
- A3 Value Chain — Visual value stream showing how business processes flow. Uses Mermaid diagrams. Helps identify where the architecture adds value.
Phase B (B*) — Business Architecture
- B1 Business Architecture — Documents the baseline (as-is) and target (to-be) business architecture, plus the gap analysis between them.
- B2 Business Capability Catalog — Lists business capabilities with current maturity levels. Helps prioritize which capabilities need architectural investment.
- B3 Business Scenarios — Detailed scenarios that drive architecture decisions. Each scenario includes actors, triggers, flows, and architecture implications.
Phase C (C*) — Information Systems
- C1 Application Architecture — Application landscape with component interactions (Mermaid diagrams), gap analysis, and target state.
- C2 Data Architecture — Data entities, data flows, ownership, and governance policies. Critical for PII handling and compliance.
- C3 Application Portfolio Catalog — Full inventory of applications (baseline + target) with lifecycle status, technology stack, and ownership.
Phase D (D*) — Technology Architecture
- D1 Technology Architecture — Infrastructure and deployment topology (Mermaid diagrams). Covers compute, networking, security zones.
- D2 Technology Standards Catalog — Approved technologies with status (Endorsed / Contained / Emerging / Retiring). Ensures alignment across teams.
Phase E (E*) — Opportunities & Solutions
- E1 Solutions Building Blocks — Maps Abstract Building Blocks (ABBs) to Solution Building Blocks (SBBs) with build-vs-buy analysis for each.
- E2 Integration Strategy — Integration patterns, API catalog, external dependencies. Defines how components talk to each other and to the outside world.
Phase F (F*) — Migration Planning
- F1 Architecture Roadmap — Phased delivery plan with work packages, dependencies, owners, and key milestones. The main planning artifact.
- F2 Migration Plan — Transition architecture states, rollback procedures, and data migration strategy. How you get from baseline to target safely.
Phase G (G*) — Implementation Governance
- G1 Compliance Assessment — Checklist ensuring implementation conforms to the architecture. Includes gate reviews and a tracker for architecture deviation requests.
- G2 Architecture Contracts — Formal agreements between the architecture team and implementation teams. Defines commitments, quality criteria, review cadence, and escalation paths.
Phase H (H*) — Change Management
- H1 Change Request Log — Formal change requests against the approved architecture. Each CR gets an impact assessment before approval.
Requirements Management (R*)
- R1 Architecture Requirements — Functional requirements, non-functional requirements, dependencies, constraints, and assumptions. Cross-referenced from all phases.
Cross-Phase (X*) — Living Documents
These are updated throughout the ADM cycle:
- X1 Decision Log — Architecture Decision Records (ADRs). Each decision captures context, options, consequences, and links to related artifacts. The most important governance artifact.
- X2 Risk & Issue Register — Risks with probability × impact scoring, mitigation plans, and an issue log. Review regularly.
- X3 Open Questions — Questions organized by category and ADM phase, with owner and target date. Drives the architecture review agenda.
- X4 Sizing Catalogue — Component sizing, capacity planning assumptions, and cost estimates. Feed this into business cases and procurement.
- X5 Traceability Matrix — End-to-end traceability from business drivers through requirements, architecture components, decisions, building blocks, and risks. The glue that holds the architecture together. Review at every gate.
YAML Front Matter
Every file includes structured metadata:
---
type: architecture-vision # Document type
togaf_phase: A # ADM phase (Preliminary, A-H, requirements-management, cross-phase)
artifact_type: deliverable # catalog | matrix | diagram | deliverable | building-block
version: 1.0.0 # Semantic version
status: draft # draft | review | approved | superseded
created: 2026-02-17
last_modified: 2026-02-17
owner: TBD # Responsible person
reviewers: [] # List of reviewers
---
The @architect agent reads this metadata to understand document maturity and ownership. Use @architect /update to modify it with full governance tracking.
Configuration
| Setting | Default | Description |
|---|---|---|
archipilot.vaultPath |
"" |
Path to the active vault folder (auto-detected if empty) |
archipilot.projectsRoot |
"architectures" |
Folder (relative to workspace root) where /new creates vaults and where discovery scans |
archipilot.contextScoping |
true |
When enabled, each command loads only the vault files relevant to its mode (e.g. /c4 loads C1–D1, /decide loads X1+A1). Disable to always load the full vault. |
archipilot.hiddenQuickActions |
[] |
List of Quick Action labels or /command shortcuts to hide from the sidebar |
Architecture
src/
extension.ts – Entry point, registration
participant.ts – @architect chat handler (all slash commands)
features.ts – ADR creation, context diagrams, vault graph
sidebar.ts – Vault Explorer, Quick Actions, Architecture Health views
vault.ts – Vault discovery, loading, switching, context windowing
prompts.ts – TOGAF system prompts per mode
updater.ts – Diff preview, file editor, audit log & backup
vault-template.ts – TOGAF vault scaffolding (28 template files)
source-scanner.ts – Workspace scanner for /scan (file discovery & content extraction)
types.ts – TypeScript interfaces
core/ – Pure logic (zero vscode imports)
archimate-exporter.ts – ArchiMate 3.2 Open Exchange XML generator
drawio-exporter.ts – Draw.io XML generator (As-Is / Target / Migration)
todo-extractor.ts – TOGAF action item scanner & prioritiser
parser.ts – Command extraction from LLM responses
validator.ts – Schema validation for architecture commands
context.ts – Context windowing & file summarisation
mutations.ts – Vault mutation helpers
index.ts – Barrel export
test/ – Unit + integration tests (152 tests)
Safety & Governance
| Feature | Description |
|---|---|
| Diff preview | Every /update shows a unified diff before applying changes |
| Vault validation | Commands checked against vault: file existence, section match, fuzzy suggestions for close names |
| Confirmation dialog | "Apply All" / "Cancel" modal before any write |
| Dry-run mode | Append --dry-run or --preview to see changes without writing |
| File backup | Pre-write copy saved to .archipilot/backups/{name}.{timestamp}.md |
| Audit log | Every mutation logged to .archipilot/audit.log (JSONL) |
| Schema validation | Commands validated for required fields, safe paths, known types |
| Confidence markers | LLM prefixes extrapolations with ⚠️ to reduce hallucination risk |
Quick-Start for Non-TOGAF Users
You don't need to be a TOGAF expert to use archipilot. Here's how to get productive in 5 minutes:
1. Scaffold a vault
@architect /new My-First-Architecture
This creates 28 template files covering all architecture domains. Don't worry about the naming — the prefixes (A1_, B1_, C1_, etc.) are just a filing convention.
2. Start with what you know
- Business goals? → Edit
A1_Architecture_Vision.md— write your project scope and drivers in plain English - Existing codebase? → Type
@architect /scan— scans your source and populates the vault automatically - Applications / systems? → Edit
C1_Application_Architecture.md— list the systems involved - Decisions to make? → Type
@architect /adr Should we use Kafka or RabbitMQ? - Risks? → Edit
X2_Risk_Issue_Register.md— add rows to the table
3. Let archipilot guide you
@architect /todo ← "What should I work on next?"
@architect /status ← "How complete is my architecture?"
@architect /audit ← "Where are the quality gaps?"
4. Evolve incrementally
You don't need to fill every file before getting value. Start with Vision → Business Scenarios → Application Architecture → Decisions. The /todo command will tell you what's missing.
Mental model
Think of the vault as a structured notebook for your architecture. Each file is a chapter. WikiLinks ([[filename]]) are cross-references. archipilot reads the whole notebook so it can answer questions, find inconsistencies, and suggest next steps.
Interactive Tutorial
archipilot includes a built-in Getting Started walkthrough inside VS Code:
- Open the Command Palette (
Cmd+Shift+P/Ctrl+Shift+P) - Type "Get Started with archipilot"
- Follow the 6 guided steps:
| Step | What you'll learn |
|---|---|
| Scan or Create | Bootstrap a vault with /scan (from code) or /new (blank template) |
| Audit | Run /audit for a unified health check (status + review + gate) |
| Analyze | Ask what-if questions with /analyze |
| Decide | Two-step decision flow — analyze first, then record with /decide |
| Update | Make governed changes with /update (unified diffs + vault validation) |
| Diagrams & Exports | Generate C4 models, ArchiMate XML, or Draw.io diagrams |
Suggested Learning Path
| Session | Goal | Commands to try |
|---|---|---|
| Day 1 | Bootstrap & explore | /scan (existing code) or /new (blank), /status, free-form questions |
| Day 2 | Capture decisions | /adr, /decide (two-step: analyze → record) |
| Day 3 | Impact & risk | /analyze, /todo |
| Day 4 | Diagrams & exports | /diagram, /graph, /c4 |
| Day 5 | Governance | /audit, /update |
| Day 6 | Enterprise tooling | /archimate, /drawio, /timeline, /scan --append |
Tip: Each session takes 15–30 minutes. By the end of the week you'll have a working architecture repository.
TOGAF Glossary
Quick reference for TOGAF terms used throughout archipilot. You don't need to memorise these — they're here when you need them.
| Term | Definition | Where it appears in archipilot |
|---|---|---|
| ADM | Architecture Development Method — TOGAF's iterative cycle (Phases A–H) for developing enterprise architecture | File prefixes (A1_, B1_, …), sidebar phase groups |
| ADR | Architecture Decision Record — a structured log entry capturing context, options, and rationale for a decision | X1_ADR_Decision_Log.md, /adr, /decide |
| ABB | Abstract Building Block — a conceptual component (e.g. "Identity Provider") | E1_Solutions_Building_Blocks.md |
| SBB | Solution Building Block — a concrete product filling an ABB (e.g. "Azure AD B2C") | E1_Solutions_Building_Blocks.md |
| ArchiMate | A modelling language for enterprise architecture (ISO/IEC/IEEE 42010 compliant) | /archimate export |
| Baseline | The current (as-is) state of the architecture | B1, C1, D1 files; Draw.io "As-Is" view |
| Target | The desired future (to-be) state | B1, C1, D1 files; Draw.io "Target" view |
| Gap Analysis | Comparison of baseline vs target to identify what must change | B1, C1, D1 files |
| Capability | A business ability (e.g. "Customer Onboarding") independent of how it's implemented | B2_Business_Capability_Catalog.md |
| Deliverable | A formally reviewed architecture work product | YAML artifact_type: deliverable |
| Phase Gate | A governance checkpoint where an ADM phase is assessed for readiness before proceeding | /gate command |
| Stakeholder | Anyone affected by or having authority over the architecture | A2_Stakeholder_Map.md |
| Traceability | End-to-end linking from business driver → requirement → component → decision → risk | X5_Traceability_Matrix.md |
| Technical Debt | Architectural shortcuts, workarounds, or deferred work requiring future resolution | X6_Technical_Debt_Log.md |
| RACI | Responsible / Accountable / Consulted / Informed matrix | A2_Stakeholder_Map.md |
| NFR | Non-Functional Requirement (performance, security, scalability, etc.) | R1_Architecture_Requirements.md |
| Work Package | A defined chunk of delivery work in the migration roadmap | F1_Architecture_Roadmap.md |
| WikiLink | Obsidian-style [[cross-reference]] between vault files |
Used everywhere for traceability |
FAQ
Getting Started
Q: Do I need an Obsidian license?
A: No. archipilot reads plain Markdown files — it doesn't require Obsidian itself. You can use any editor. The vault structure and [[WikiLinks]] follow Obsidian conventions, but the files are standard .md.
Q: Do I need to know TOGAF? A: No. See the Quick-Start for Non-TOGAF Users section. The templates provide structure; archipilot guides you on what to fill in next.
Q: What's the minimum vault to start getting value?
A: A vault with just A1_Architecture_Vision.md and X1_ADR_Decision_Log.md is enough for free-form Q&A and decision tracking. Use /new to scaffold a full set of 27 files if you want the complete framework.
Using Commands
Q: What's the difference between /analyze and /audit?
A: /analyze answers a specific what-if question ("What if we drop the API Gateway?"). /audit performs a comprehensive health check of the entire vault — combining status dashboard, quality review, and phase gate assessment. Use --quick for status + review only, or --full for the complete check. (/review and /gate still work as aliases.)
Q: Can I undo a /update?
A: Yes. Every write creates a backup in .archipilot/backups/ with a timestamp. You can also use --dry-run to preview before applying. The audit log (.archipilot/audit.log) records every mutation.
Q: Why does /status show "Draft" for files I've edited?
A: The status comes from the YAML front matter status: field. Use @architect /update Bump A1_Architecture_Vision status to review to update it through governance.
Q: Can I use archipilot without GitHub Copilot? A: No — archipilot is a Copilot Chat Participant and requires an active GitHub Copilot subscription. The LLM calls go through Copilot's API.
Exports & Diagrams
Q: Which tools can open the ArchiMate export? A: Archi (free, open-source), BiZZdesign, Sparx EA, ADOIT, and any tool supporting the ArchiMate 3.2 Open Exchange Format.
Q: The Mermaid diagrams don't render in my editor. What do I need?
A: Install the Markdown Mermaid VS Code extension. For .mermaid files, install Mermaid Preview.
Q: Can I export to PowerPoint or Confluence?
A: Not directly. Export to Draw.io (.drawio) and then use Draw.io's built-in PowerPoint/Confluence export, or export ArchiMate XML and import into your enterprise tooling.
Vault Management
Q: Can multiple people work on the same vault?
A: Yes — the vault is just a folder of Markdown files. Use Git for version control, branches, and merge. The YAML front matter tracks last_modified and owner for coordination.
Q: How do I archive a completed architecture?
A: Set all document statuses to approved, run /audit --full for a final quality check, then commit the vault to Git with a release tag. You can keep it in the workspace as a read-only reference.
Q: What happens if I rename or move vault files?
A: WikiLinks will break. Run /todo or check the Architecture Health sidebar — it detects broken links. Use /update to fix cross-references after renaming.
Q: Can I add custom files to the vault?
A: Yes. Any .md file in the vault folder is included in context. Follow the prefix convention (e.g. C4_API_Catalog.md) so the sidebar groups it correctly, and add a WikiLink from 00_Architecture_Repository.md.
Contributing
Contributions are welcome! See CONTRIBUTING.md for guidelines.
Please read our Code of Conduct before participating.
Changelog
See CHANGELOG.md for release history.
License
This project is licensed under the MIT License — see the LICENSE file for details.