busirocket-nextjs-route-handlers

The repo is self-discoverable so agents and tools can see what skills exist

without cloning and running scripts:

• [llms.txt](llms.txt) – Human- and machine-readable index at repo root.

Lists each skill with name, description, and path to SKILL.md. Regenerate

with yarn generate:llms-txt after adding or changing skills.

• [available_skills.xml](available_skills.xml) – XML

for

[injecting into agent system prompts](https://agentskills.io/integrate-skills).

Regenerate with yarn to-prompt:file. For correct paths on your

machine, run yarn to-prompt after cloning (or use the installed skill paths

from npx skills add).

Both files are committed so consumers can fetch them from the repo (e.g. raw

GitHub URL) for discovery; the

[Agent Skills specification](https://agentskills.io/specification) describes

discovery, metadata loading, and activation.

Core Skills (Auto-activated)

• busirocket-core-conventions - General engineering conventions, file

discipline, boundaries, naming/layout, and anti-patterns

• busirocket-typescript-standards - TypeScript export discipline, type

conventions, Next.js special-file exceptions

• busirocket-react - React component and hook structure rules

(one-component/one-hook per file) plus Zustand state management

• busirocket-nextjs - Next.js App Router patterns, thin route handlers,

validation, response shapes

• busirocket-validation - Validation strategy (Zod schemas, guard helpers)

at boundaries

• busirocket-refactor-workflow - Strict refactoring workflow with quality

gates

• busirocket-tailwindcss-v4 - Tailwind CSS v4 setup and styling strategy
• busirocket-rust - Rust language and module standards

(one-thing-per-file, SQL/prompt separation, boundaries)

• busirocket-tauri - Tauri-specific layout and commands checklist (desktop

apps)

Optional Skills (Manual invocation only)

• busirocket-supabase - Supabase access patterns and service boundaries

(use /busirocket-supabase when working with Supabase)

Recommended: Using npx skills add

The easiest way to install these skills is using the official installer, which

automatically detects and installs to all compatible agents:

```bash

npx skills add BusiRocket/agents-skills

```

Or use the deprecated (but still working) command:

```bash

npx add-skill BusiRocket/agents-skills

```

The installer will:

• Detect all compatible agents on your system (Cursor, Claude Code, Codex,

GitHub Copilot, etc.)

• Let you select which skills to install
• Install them globally or per-project
• Create symlinks automatically

Skills are automatically activated by agents when relevant tasks are detected.

Optional skills can be invoked manually:

```

/busirocket-supabase

```

Zustand state management patterns are included in busirocket-react; use

/busirocket-react when working with Zustand.

Prerequisites

• Node.js 18+
• npm or yarn

Setup

```bash

npm install

# or

yarn install

```

Scripts

• yarn build - Full pipeline: format, version check, validate, generate

llms.txt, generate available_skills.xml. Run before release or when

preparing the repo.

• npm run format - Format all Markdown files with Prettier
• npm run format:check - Check Markdown formatting without writing

(CI-friendly)

• yarn generate:llms-txt - Regenerate [llms.txt](llms.txt) from skill

frontmatter (Node-only). Run after adding or editing skills.

• yarn to-prompt - Generate XML for all skills (stdout).

Redirect to a file or pipe as needed. See

[Integrate skills into your agent](https://agentskills.io/integrate-skills)

for how to use the output in agent system prompts. Uses the same venv as

yarn validate if present.

• yarn to-prompt:file - Generate and write

[available_skills.xml](available_skills.xml) at repo root (for discovery and

CI). Uses same venv as yarn validate.

• yarn validate - Validate all skills with

[skills-ref](https://github.com/agentskills/agentskills/tree/main/skills-ref)

(frontmatter, naming). Uses the project venv if present (see below).

• yarn validate:install - Create a Python venv at .venv-validate and install

skills-ref. Run once; then yarn validate and yarn to-prompt use it.

Requires Python 3.11+ on PATH.

• yarn validate:uninstall - Remove the .venv-validate venv.
• yarn version:check - Ensure package.json version matches

metadata.version in each skill's SKILL.md frontmatter. Fails if any skill is

out of sync (used by yarn build).

Each skill contains:

• SKILL.md - Main instructions with frontmatter metadata
• rules/ - Rule files loaded on demand (progressive disclosure)

This repo follows

[Claude skill authoring best practices](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/best-practices).

Checklist for maintainers:

• [ ] Descriptions in third person (e.g. "Applies…", "Enforces…")
• [ ] References one level deep (SKILL.md → rules only; no rule → rule links)
• [ ] Forward slashes in paths (no Windows-style \)
• [ ] Table of contents in rule files longer than 100 lines
• [ ] No time-sensitive content (dates, "before/after X version")
• [ ] Consistent terminology (route handler, services/, utils/, types/)

MIT