Skip to content

mmaideveloper/aiarchitecttoolkit / skills/generate-add

generate-add

Verified

Create, update, and review a source-backed Architecture Design Document (ADD) from reviewed use cases, applicable Business Decision Records, repository evidence, constraints, quality attributes, C4 views, and ADRs. Use when a change needs an overall technical design, an ADD reference must be revised, or implementation readiness must be assessed without writing code.

FileKindSize
SKILL.mdskill md2.0 KB
agents/openai.yamlconfig233 B
assets/add-template.mdmarkdown3.3 KB
references/add-quality.mdmarkdown687 B
---
name: generate-add
description: Create, update, and review a source-backed Architecture Design Document (ADD) from reviewed use cases, applicable Business Decision Records, repository evidence, constraints, quality attributes, C4 views, and ADRs. Use when a change needs an overall technical design, an ADD reference must be revised, or implementation readiness must be assessed without writing code.
---

# Generate ADD

## Workflow

1. Read `AGENTS.md`, the project profile, reviewed use cases, applicable BDRs, existing architecture, relevant code, deployment sources, diagrams, and decisions.
2. Resolve the next `ADD-NNN` or the referenced ADD. Never reuse or renumber identifiers.
3. Copy [assets/add-template.md](assets/add-template.md) and apply [references/add-quality.md](references/add-quality.md).
4. Define scope, drivers, constraints, system context, responsibilities, data flows, interfaces, deployment, operations, risks, and migration/rollback.
5. Link C4 views rather than embedding several abstraction levels in one diagram. Request `$generate-c4` only for views that answer a concrete stakeholder question.
6. Identify independently changeable choices as ADR candidates. Link proposed ADRs; treat only accepted ADRs as constraints.
7. Run the privacy, security, clinical/legal safety, and human-oversight assessment. Stop on an unresolved compliance conflict.
8. Save under `architecture/design/ADD-NNN-<slug>.md` unless an established layout exists.

## Rules

- Do not invent owners, technologies, protocols, service levels, classifications, approvals, or data flows.
- Distinguish current, transition, target, and conceptual states.
- Include failure modes and operational ownership, not only the happy path.
- Keep real personal/health data, credentials, secrets, and sensitive endpoints out of the document.
- Keep status `Draft` or `In Review` until authoritative approval is cited.

## Completion

Report the document path, linked use cases, required C4 views, ADR candidates, compliance blockers, open questions, and implementation-readiness result.
brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).Stars: 322kUpdated Claude CodeCodexVerifiedInstall
brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).Stars: 198kUpdated Claude CodeCodexVerifiedInstall