agentdev-architecture-advisory

Reviews requirement drafts for architecture conflicts before Decision judgment in req-define workflows.

Updated Mar 7, 2026
One-click install
npx skills add https://github.com/yogata/agent-dev-flow --skill agentdev-architecture-advisory-yogata
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agentdev-architecture-advisory
Source: https://github.com/yogata/agent-dev-flow/tree/main/src/opencode/skills/agentdev-architecture-advisory
Command: npx skills add https://github.com/yogata/agent-dev-flow --skill agentdev-architecture-advisory-yogata

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? When defining requirements, changes may conflict with existing REQ/Decision/Design documents or cross responsibility boundaries of commands, skills, and workflows. This Skill provides a pre-confirmation architecture review step so requirement drafts are checked for architectural impact before Decisions are finalized. ## Core Features & Use Cases - Standard Consultation Template: Structures advisory requests with five elements: requirement candidates, conflict candidates, Decision candidates, existing Decision relations, and specific judgment questions. - Four-Label Output Contract: Classifies advisory output as confirmed facts, assumptions, user-confirmation items, or blockers, with final classification authority retained by the parent agent. - Use Case: When a new requirement may alter command/skill responsibility boundaries or overlap an existing Decision, req-define consults the architecture advisory subagent before Step 5 Decision judgment, preventing unverified assumptions from entering the requirement draft. ## Quick Start Ask the agent to run an architecture impact review on the current requirement draft before finalizing Decision judgments in req-define.

Frequently Asked Questions about agentdev-architecture-advisory

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check architecture impact before finalizing a requirement?▼

Invoke the architecture advisory review after requirement expansion (Step 4) and before Decision judgment (Step 5) in req-define. The parent agent builds a five-element consultation template covering requirement candidates, conflicts, Decision candidates, existing Decision relations, and judgment questions.

When should architecture advisory review be triggered in req-define?▼

Trigger it when requirements may conflict with existing REQ/Decision/Design documents, create or update Decisions, change command/skill/workflow responsibility boundaries, affect external execution agents, or span multiple commands with technically uncertain tradeoffs.

What are the four output labels in architecture advisory review?▼

The labels are confirmed facts (backed by documents or user agreement), assumptions (unverified inferences), user-confirmation items (requiring user consent), and blockers (unresolvable open issues). The parent agent holds final classification authority over these labels.

Can the architecture advisory subagent edit requirement files directly?▼

No. The advisory subagent only reads and advises; it never edits drafts, REQ, Decision, Design, Issues, or PRs. The parent agent classifies the advice and reflects confirmed items into the draft.

What happens if the architecture advisory agent is unavailable?▼

If advisory conditions are met but the advisory agent is unavailable, req-define reports a blocker instead of silently skipping the review. This prevents unreviewed architectural impacts from entering finalized requirements.