Agent skill · Databases

architect-agent-skill

Technical workflow for architect-agent - PRD creation with mandatory database context, entity design with Zod, iteration review, and agent coordination in the iterative v2.0 system

majiayu000github.com/majiayu000GitHub ↗
claude-codeMIT
Install
npx skills add majiayu000/claude-skill-registry --skill architect-agent-skill-fercracix33-poli2-clean-architec --agent claude-code

Same command for any agent — swap --agent for codex, cursor, copilot.

Facts
Files in the skill folder: 2
SKILL.md size: 23 KB
Bundled scripts: none
Path: skills/agent/architect-agent-skill-fercracix33-poli2-clean-architec/SKILL.md
Open the folder on GitHub →
Where it comes from
Stars: 534
Language: HTML

Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.

Review
written from the skill's own SKILL.md · Aug 5, 2026

What it does

The skill prescribes a complete, phase-driven workflow for an architect role to produce a master PRD and data contracts. It enforces mandatory Phase 0 research against a Supabase database, Phase 1 clarification dialogue, Phase 2 context research for design patterns, Phase 3 generation of a 14-section master PRD with a critical Data Contracts section, Phase 4 scaffolding of a feature directory and implementation of an entities.ts contract (no business logic), Phase 5 creation of agent-specific request documents for multiple agent roles, and Phase 6 review iterations before progression. It provides concrete TypeScript/Zod templates and SQL examples for RLS, and outlines required artifacts and validation checks.

How it works

  • Phase 0 (Mandatory) requires querying existing database context via mcp__supabase__* helper calls to list tables, inspect schemas, review FKs, and inspect RLS policies. Output should identify multi-tenancy patterns, naming conventions, RLS patterns, FK patterns, and duplication risks.
  • Phase 1 provides a structured questionnaire to gather permissions, authorization, functional, data, side effects, and performance requirements, encouraging context-rich questions.
  • Phase 2 mandates context7 research calls for Zod patterns, API patterns, and RLS practices to inform PRD and entity design.
  • Phase 3 requires generating a Master PRD with 14 sections, emphasizing a Data Contracts section containing specific Zod schemas and TypeScript types. It includes optional CASL sections if authorization is required and a detailed RLS policy template.
  • Phase 4 creates a feature directory structure and demands implementing entities.ts only, using a provided template that defines Zod schemas for a main entity, create/update variants, and TS types, with no business logic.
  • Phase 5 instructs producing agent-request.md files for Test Agent, Implementer, Supabase Agent, and UI/UX Expert, each populated with context, objectives, and detailed requirements.
  • Phase 6 enforces review of every iteration before proceeding.

When to use it

Use when designing a new architect-focused feature requiring a robust PRD, multi-tenant database considerations, Zod-based data contracts, and coordinated agent-driven development in an iterative system. Activate Phase 0 before any questions or design work and proceed sequentially through phases to ensure mandatory checks and artifacts are produced.

What it can touch

  • Supabase database context via mcp__supabase__* calls (Phase 0).
  • Zod schemas and TypeScript types (Phase 2 guidance; Phase 4 implementation).
  • PRD artifacts in Markdown under PRDs directory (Phase 3).
  • File system scaffolding for feature directories and API routes (Phase 4).
  • Agent request documents for multiple agent roles (Phase 5).

Caveats

  • This skill requires exact adherence to the multi-phase process; skipping phases is not supported.
  • The entities.ts implementation is solely a contract (data structures) with no business logic.
  • All code and templates must be used exactly as shown; no speculative implementations beyond the specified patterns.
  • The skill references external templates and repository conventions (e.g., PRDs templates and mcp context queries) that must exist in the environment to be operable.
From the SKILL.md

# Architect Agent Skill Complete technical workflow and best practices for the Chief Architect role. --- ## 6-PHASE WORKFLOW ### PHASE 0: Pre-Discovery Research (MANDATORY) **⚠️ CRITICAL**: Query existing database context BEFORE asking user any questions. **Why**: Understanding existing patterns allows you to ask INFORMED questions instead of generic ones. #### Supabase MCP Research (REQUIRED) ```typescript // 1. List all existing tables mcp__supabase__list_tables({ schemas: ['public'] }) // 2. Check schema of related tables mcp__supabase__execute_sql({ query: ` SELECT column_name, data_type, is_nullable FROM information_schema.columns WHERE table_name = 'existing_table' ORDER BY ordinal_position; ` }) // 3. Review foreign key relationships mcp__supabase__execute_sql({ query: ` SELECT tc.table_name, kcu.column_name, ccu.table_name AS foreign_table_name FROM information_schema.table_constraints AS tc JOIN information_schema.key_column_usage AS kcu ON tc.constraint_name = kcu.constraint_name JOIN information_schema.constraint_column_usage AS ccu ON ccu.constraint_name = tc.constraint_name WHERE tc.constraint_type = 'FOREIGN KEY' AND tc.table_schema = 'public'; ` }) // 4. Review exist

What's inside
Steps it walks through
  1. 6-PHASE WORKFLOW
  2. PHASE 0: Pre-Discovery Research (MANDATORY)
  3. PHASE 1: Clarification Dialogue
  4. PHASE 2: Context7 Research (MANDATORY)
  5. PHASE 3: Generate Master PRD
  6. PHASE 4: Create Directory Structure & Entities
  7. PHASE 5: Write Agent Request Documents
  8. PHASE 6: Review Iterations & Approve/Reject
  9. OPTIONAL: Enable Parallelism via Handoffs
  10. PRE-HANDOFF VALIDATION CHECKLIST
  11. Context Research Completed
  12. PRD Completeness
  13. Directory Structure
  14. Entities Implementation
Ships with 1 file
  • metadata.json
Commands it runs
cp PRDs/_templates/00-master-prd-template.md PRDs/{domain}/{number}-{feature}/architect/00-master-prd.md
Main feature directory
mkdir -p app/src/features/{feature-name}
Subdirectories
mkdir -p app/src/features/{feature-name}/components
mkdir -p app/src/features/{feature-name}/use-cases
mkdir -p app/src/features/{feature-name}/services
Create entities.ts (YOU implement this)
touch app/src/features/{feature-name}/entities.ts
Create placeholder files for other agents
More from claude-skill-registry
All skills →
About this skill
What does the architect-agent-skill skill do?

Technical workflow for architect-agent - PRD creation with mandatory database context, entity design with Zod, iteration review, and agent coordination in the iterative v2.0 system

How do I install it?

Run `npx skills add majiayu000/claude-skill-registry --skill architect-agent-skill-fercracix33-poli2-clean-architec --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.

Where does this skill come from?

From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.

Is a popular skill a good skill?

Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.

Keep going