Claude Skills by gurenjs

Development workflow commands for the Guren monorepo. Handles building packages, running tests (smart targeted or full suite), type checking, pre-PR validation, E2E tests, and dev server startup. Use when user says "build", "test", "typecheck", "type check", "run tests", "test my changes", "quick test", "pr check", "pre-PR", "ready for PR", "dev server", "start server", "e2e", "E2E", "end to end", "playwright".

Draft an RFC for a significant change to the Guren framework itself, per contributing/rfc-process.md. Use when the user says "write an RFC", "create an RFC", "propose a breaking change", "RFC for X", or is about to design a new public API, cross-cutting feature, or major architecture decision for this monorepo.

Install or refresh the Guren agent harness (rules, skills, hooks, MCP config) inside an existing Guren app for Claude Code, Codex, Cursor, Copilot, and OpenCode, and explain the guren context / check / audit loop. Use when the user has a Guren app and says "set up the harness", "agent:init", "sync the harness", "install Guren rules for Cursor", or asks how to work in a Guren app with an AI agent. For a directory with no Guren app, use guren-new-app instead.

Scaffold a new Guren application (Bun + Hono + Drizzle + Inertia, Laravel-shaped) in a directory that has no Guren app yet, and hand off to the app's own agent harness. Use when the user wants to start a Guren project, says "new Guren app", "create a Guren project", "scaffold Guren", or asks what Guren is before choosing a stack. For a directory that already has a Guren app, use guren-harness instead.

Expose an existing route to AI agents as an MCP tool via route `.agent()` metadata. Use when the user asks to "expose this to agents", "make an MCP tool", "agent tool", "agent interface", "let Claude call this endpoint", or asks which of the app's routes agents can already reach.

Write, wire and test an in-process AI agent with @guren/plugin-ai (an Agent class that calls a language model and reaches the app through its .agent() routes). Use when the user asks to "add an AI feature", "call Claude/OpenAI from the app", "summarize/classify/triage with an LLM", "make an AI agent", "add a chat", or mentions app/Ai/Agents, config/ai.ts, appTools(), fakeAi() or make:ai-agent. Not for durable Workers agents (make:agent) or for exposing routes to external agents (agent-interfa...

Database management for Guren with safety checks. Handles migrations, rollbacks, seeding, resets, and container management. Use when user mentions "database", "migration", "migrate", "rollback", "seed", "reset", "db:up", "db:down", "db:reset", "fresh database", or database-related tasks.

Development workflow commands for a Guren application. Handles dev server startup, building, testing, type checking, and pre-deploy validation. Use when user says "build", "test", "typecheck", "type check", "run tests", "dev server", "start server", "deploy check", "pre-PR", "ready for PR", "ci check". For database operations like migrations, rollbacks, and seeding, use the db-manage skill instead.

Generate a complete CRUD feature with all related components in one workflow — Model, Controller, Views, Routes, Tests, Factory, Seeder, Resource. Use when user wants to build out an entire entity at once. Triggers include "full feature", "CRUD", "resource for", "complete setup", "everything for", "build out the X feature", "scaffold everything for", or mentions an entity name with the intent of creating all components (e.g., "I need a Product entity"). For creating a single component, use th...

Work an issue the way this app tracks work. GitHub Issues and Projects hold the task, its progress and its assignee; docs/ holds the decision and links to the issue. Use when the user says "work on #412", "pick up the issue", "open an issue for this", "update the board", "move it to In Progress", "what is in flight on User", or hands you a GitHub issue URL.

Human-browsable Guren API reference — code patterns and examples for every subsystem (Controllers, Models, Routes, Middleware, Authentication, Authorization, Events, Jobs, Queue, Mail, Cache, Validation, Broadcasting, Notifications, Storage, Scheduling, I18n, Encryption, Health Checks, Error Handling, Container/ServiceProvider, Console Commands, API Resources). Use when the user asks "how to", "how does", "example of", or "what is" about a Guren API. Agent-critical signatures are already avai...

Implement an approved implementation plan (a `*.plan.json`, RFC 0030) one derived step at a time — `plan:next` names the step and what it covers, you implement exactly that, `plan:verify` records the result, one commit per step. Use when the user says "implement the plan", "next step", "continue the plan", "work through the plan", or names a plan file with the intent of building it.

Install or author Guren plugins — npm packages that extend a Guren app via the ServiceProvider system. Use when the user asks to "install a plugin", "add @guren/plugin-*", "create a plugin", "build a Guren plugin", "make:plugin", or wants a reusable package that registers services, middleware, or CLI commands into a Guren app.

Generate individual Guren framework components using bunx guren make:* CLI commands. Creates a single component at a time — controllers, models, views, middleware, jobs, events, listeners, mail, tests, resources, validators, factories, seeders, providers, commands, exceptions, channels, notifications, migrations, routes. Use when user says "create a controller", "generate a model", "scaffold middleware", "make a job", "add a job", "new event", "make middleware", "make:controller", or asks for...