Phase 5 adr

Skill exchanet/method_enterprise_builder_planning/agents/antigravity/.agent/skills/phase-5-adr

🏗️ Universal 8-phase methodology for planning and building enterprise-grade, mission-critical software with AI coding agents. Supports Cursor AI, Claude Code, Kimi Code, Windsurf & Google Antigravity. Includes ADR validator, microtask linter & CI/CD templates. ≥99% test coverage. MIT.

Install

npx -y skills add exchanet/method_enterprise_builder_planning --skill phase-5-adr

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Phase 5 — Architecture Decisions with ADR generation and validation

SKILL.md

3.1 KB, as published. Nobody here has run it

Skill: Phase 5 — Architecture Decisions (ADR)

Purpose

Execute Phase 5: apply the architecture decision tree, create ADRs for every significant decision, generate C4 diagrams, and validate all ADRs before marking as Accepted.

When to use

Phase 5 of the full 8-phase cycle
When the user needs architecture guidance independently
When designing a new module or refactoring an existing system

Decision Tree

Answer these questions in order to select the primary architectural pattern:

Single team or multiple autonomous teams?
- Single → Modular Monolith
- Multiple → Microservices
Does business complexity require a rich domain model?
- Yes → Hexagonal Architecture / DDD
- No → Layered Architecture
Is full state audit trail required?
- Yes → Event Sourcing
- No → Standard persistence
Are reads and writes radically different in volume?
- Yes → CQRS
- No → Standard CRUD
Are there distributed transactions across services?
- Yes → Saga pattern (Orchestration or Choreography)
- No → Local transactions

ADR Template (mandatory sections)

Every ADR must include:

# ADR-NNN: Short description of the decision

**Date:** YYYY-MM-DD
**Status:** Proposed | Accepted | Deprecated | Superseded
**Deciders:** Name, Role

## Context

[2-3 paragraphs: problem, constraints, why decide now]

## Decision

[Clear statement of the decision]

## Alternatives Considered

| Alternative | Pros | Cons | Why rejected |
|---|---|---|---|
| Option A | ... | ... | Rejected because [reason] |
| Option B | ... | ... | Rejected because [reason] |

## Consequences

**Positive:**
- [benefit 1]

**Negative:**
- [drawback 1] — mitigated by [mitigation]

## Compliance Impact

[State which standards affected, or confirm no impact]

Validation step

After generating each ADR, validate it:

node packs/enterprise-architecture-pack/validators/adr-validator/src/index.ts \
  --format=console docs/adr/ADR-NNN-title.md

Or in Antigravity Manager mode, use code execution:

import subprocess
subprocess.run(['node', '--experimental-strip-types', 
  'packs/enterprise-architecture-pack/validators/adr-validator/src/index.ts',
  '--format=console', 'docs/adr/ADR-001-isolation.md'])

Gate: If validation fails (any error-severity rule), fix the ADR and re-run. Only mark as Accepted after validation passes.

C4 Diagrams

Generate minimum:

C4 L1 (System Context): External actors and systems
C4 L2 (Container): Services, databases, message queues

Use Mermaid format (default) or PlantUML if requested.

Output artifacts

docs/adr/ADR-001-*.md, ADR-002-*.md, ... (one per decision)
C4 diagrams (inline in ADR or separate files)
Artifact type in Manager mode: adr_document