agent-creator
Use when creating expert agents. Generates agent.md with frontmatter, hooks, required sections, and skill references.
Use when applying SOLID principles to Next.js architecture — modular structure, file size limits, interface separation, JSDoc.
$ npx -y skills add fusengine/agents --skill solid-nextjs --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/solid-nextjsContext preview
The summary Claude sees to decide when to auto-load this skill.
Use when applying SOLID principles to Next.js architecture — modular structure, file size limits, interface separation, JSDoc.
name: solid-nextjs description: Use when applying SOLID principles to Next.js architecture — modular structure, file size limits, interface separation, JSDoc. version: 1.0.0 user-invocable: true references: references/solid-principles.md, references/single-responsibility.md, references/open-closed.md, references/liskov-substitution.md, references/interface-segregation.md, references/dependency-inversion.md, references/architecture-patterns.md, references/templates/server-component.md, references/templates/client-component.md, references/templates/service.md, references/templates/hook.md, references/templates/interface.md, references/templates/store.md, references/templates/action.md, references/templates/api-route.md, references/templates/validator.md, references/templates/factory.md, references/templates/adapter.md, references/templates/error.md, references/templates/test.md, references/templates/middleware.md, references/templates/prisma.md, references/templates/i18n.md, references/templates/query.md
<objective> Enforces SOLID and DRY architecture on Next.js 16 projects: files under 100 lines (pages < 50, Server Components < 80, Client Components < 60, Server Actions < 30), types isolated in `interfaces/` directories (never in components/pages), mandatory JSDoc on exports, and the module/`modules/cores/` directory structure. Requires researching current Next.js/React/Prisma docs (via Context7/Exa with the current year) before writing any code, and grepping for existing utilities before creating new ones.
Provides a dedicated reference guide per SOLID letter (SRP file-splitting, OCP composition/plugin patterns, LSP contract compliance, ISP role-based interfaces, DIP constructor injection/factories) plus 23 ready-to-copy templates covering components, hooks, services, stores, Server Actions, API routes, validators, Prisma singletons, i18n, and queries. This is a cross-cutting code-quality skill, not a feature skill — for the framework APIs themselves see nextjs-16 and its sibling skills. </objective>
**Today: January 2026** - ALWAYS use the current year for your searches. Search with "2025" or "2026", NEVER with past years.
**CRITICAL: Check today's date first, then search documentation and web BEFORE writing any code.**
1. **Use Context7** to query Next.js/React official documentation 2. **Use Exa web search** with current year for latest trends 3. **Check Vercel Blog** of current year for new features 4. **Verify package versions** for Next.js 16 compatibility
**Search queries (replace YYYY with current year):**
---
**Before ANY implementation:** 1. Explore project structure to understand architecture 2. Read existing related files to follow established patterns 3. Identify naming conventions, coding style, and patterns used 4. Understand data flow and dependencies
**Before writing ANY new code:** 1. **Grep the codebase** for similar function names, patterns, or logic 2. Check shared locations: `modules/cores/lib/`, `modules/cores/components/`, `modules/cores/hooks/` 3. If similar code exists → extend/reuse instead of duplicate 4. If code will be used by 2+ features → create it in `modules/cores/` directly 5. Extract repeated logic (3+ occurrences) into shared helpers 6. Run `npx jscpd ./src --threshold 3` after creating new files
---
Before ANY implementation, spawn 3 agents in parallel, one `Agent` call each with a `name`:
1. **fuse-ai-pilot:explore-codebase** - Analyze project structure and existing patterns 2. **fuse-ai-pilot:research-expert** - Verify latest docs for all stack technologies 3. **mcp__context7__query-docs** - Check integration compatibility
After implementation, run **fuse-ai-pilot:sniper** for validation.
---
See `references/architecture-patterns.md` for complete structure with feature modules and cores directory.
/** * Fetch user by ID from database. * * @param id - User unique identifier * @returns User object or null if not found * @throws DatabaseError on connection failure */ export async function getUserById(id: string): Promise<User | null>
modules/auth/src/ ├── interfaces/ # Types ONLY │ ├── user.interface.ts │ └── session.interface.ts ├── services/ # NO types here └── components/ # NO types here
---
Each SOLID principle has a dedicated reference guide:
1. **`references/single-responsibility.md`** - One class/function = one reason to change
2. **`references/open-closed.md`** - Extend via composition, not modification
3. **`references/liskov-substitution.md`** - Contract compliance & behavioral subtyping
4. **`references/interface-segregation.md`** - Many focused interfaces beat one fat interface
5. **`references/dependency-inversion
A plugin ecosystem that turns Claude Code into a supervised, multi-agent development environment.
Repo: fusengine/agents
Use when creating expert agents. Generates agent.md with frontmatter, hooks, required sections, and skill references.
Use when starting ANY development task -- feature, bug fix, refactor, hotfix (triggers: implement, create, build, fix, add feature, refactor, develop).
Use when creating a feature/component or adding functionality. Fires BEFORE APEX Analyze to refine requirements via structured questioning.
Use before a root-cause, done/verified claim, irreversible action, or 2nd-time fix reaches the owner (APEX or plain conversation); also fires at every…
Use when validating code quality after modifications -- SOLID compliance, DRY duplication, linter errors, architecture violations. Do NOT use for functional…
Use when an expert agent self-reviews and self-corrects code after the Execute phase, before sniper validation (BMAD-METHOD elicitation techniques).