controller
Skill4 starsChanged 2 months agoMITView on GitHub
Create an HTTP controller (API endpoint) with Zod schemas and validation. Use when adding endpoints like POST /products, GET /users/:id. Use this skill whenever implementing REST API routes, defining request/response schemas, or adding new HTTP endpoints to any bounded context. Dispatch hub — routes to typescript/go variants by file extension.
Files
Controller
HTTP endpoint. Thin shell: validate request via schema, call a use case, return the response. Never thinks, never touches repositories, never owns business rules.
The shape and idioms of this primitive differ per backend language; the lang-specific playbooks below carry the concrete rules, code, and bad practices.
Cross-language principle — query/list filters live on the controller, not in contracts
A filter/toggle/sortBy/groupBy literal set that only narrows one list endpoint is a UI concern of that controller. Define it inline on the controller's query schema (the SDK still generates a type-safe enum from this endpoint's OpenAPI). Promote a value to packages/contracts only when another service (Go ↔ TS) or a persisted DB column must agree on it — that is the sole reason the cross-language source of truth exists. A single-screen filter never qualifies. See each lang playbook ("Query/List Filters Are Controller-Local") for the worked example.
Language variants
| Backend | Playbook | Registry |
TypeScript — packages/api/typescript/ | typescript/SKILL.md | typescript/registry.yaml |
Go — packages/api/go/ | go/SKILL.md | go/registry.yaml |
How dispatch works
scripts/review.ts infers the language from the file extension (.ts / .go) and loads the matching <lang>/registry.yaml. The CLI scaffolder takes a --lang flag (--lang=typescript / --lang=go) or infers it from cwd. When editing or reviewing code, open the lang playbook matching the file you're touching — patterns and bad practices are coded per-language (e.g. CTRL-GO-01).
In this repository45 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- agentSkill
- bounded-contextSkill
- clean-branchSkill
- commitSkill
- componentSkill
- db-modellingSkill
- ddd-modelingSkill
- ddd-specSkill
- design-systemSkill
- desktop-shellSkill
- e2eSkill
- entitySkill
- enumSkill
- errorsSkill
- eventSkill
- formSkill
- grill-with-docsSkill
- handlerSkill
- handoffSkill
- middlewareSkill
- migrateSkill
- prdSkill
- primitiveSkill
- projectionSkill
- projectorSkill
- querySkill
- repositorySkill
- reviewSkill
- routeSkill
- saas-researchSkill
- schedulerSkill
- schemaSkill
- sdkSkill
- serviceSkill
- spec-reviewSkill
- storeSkill
- storybookSkill
- task-breakdownSkill
- testSkill
- trace-analysisSkill
- ui-compositionSkill
- usecaseSkill
- user-storiesSkill
- value-objectSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.