bmad-agent-architect

Adopts a system architect persona to produce technical architecture decisions from product requirements.

Updated Mar 1, 2026
One-click install
npx skills add https://github.com/Hexalith/Hexalith.Parties --skill bmad-agent-architect-hexalith
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bmad-agent-architect
Source: https://github.com/Hexalith/Hexalith.Parties/tree/main/.agents/skills/bmad-agent-architect
Command: npx skills add https://github.com/Hexalith/Hexalith.Parties --skill bmad-agent-architect-hexalith

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams moving from product requirements and UX designs into implementation often lack a consistent architectural voice, leading to ad-hoc technical decisions. This Skill activates Winston, a System Architect persona that converts PRDs and UX into pragmatic architecture decisions and keeps implementation aligned during the BMad Method solutioning phase. ## Core Features & Use Cases - Persona Activation: Loads a customizable architect identity (role, communication style, principles) from layered TOML configuration files with base, team, and user overrides. - Architecture Menu: Dispatches capabilities such as producing the architecture spine (bmad-architecture) and checking implementation readiness before sprint planning (bmad-sprint-planning). - Persistent Context: Loads persistent facts, project config, and planning artifact locations so architectural guidance stays consistent across the session. - Use Case: After finishing a PRD, a team lead asks Winston to define the architecture spine, then verifies planning artifacts are complete before opening sprint planning. ## Quick Start Ask the agent to act as Winston the system architect and produce the architecture spine for your current PRD.

Frequently Asked Questions about bmad-agent-architect

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

FAQPage Schema
How do I customize the architect agent persona?

Customize the persona by editing TOML files in three layers: the skill's customize.toml defaults, team overrides in _bmad/custom, and personal user overrides. Scalars override, tables deep-merge, and arrays append according to the BMad structural merge rules.

What does the system architect agent do after activation?

After activation it resolves the agent configuration, adopts the Winston persona, loads persistent facts and project config, greets the user, then presents a numbered menu of capabilities. You can dispatch items by number, code, or a fuzzy description match.

Can the architect agent check implementation readiness before sprint planning?

Yes, the IR menu item invokes the bmad-sprint-planning skill to verify planning artifacts are complete and aligned. It opens sprint planning and can stop after the readiness gate or continue into tracking.

What happens if the customization resolver script fails?

If the resolve_customization.py script fails, the agent falls back to manually reading the three TOML files in base, team, then user order. It applies the same structural merge rules itself, skipping any missing files.

When should I use this architect agent instead of writing architecture docs directly?

Use it during the BMad solutioning phase when you need consistent architectural decisions tied to a PRD and UX artifacts. It is less suitable for one-off diagrams or teams not following the BMad Method artifact structure.