Middleware

Use beforeAction and afterAction hooks for policy, audit, and shaping output.

Middleware

beforeAction

Use beforeAction for authorization checks, audit prep, or guard clauses.

app/actions/todos.ts

import { defineAction } from "@zapaction/core";
import { z } from "zod";

type AppContext = { userId: string; role: "admin" | "user" };

export const archiveTodo = defineAction<
  { id: string },
  { ok: true },
  AppContext
>({
  input: z.object({ id: z.string() }),
  output: z.object({ ok: z.literal(true) }),
  beforeAction: async ({ ctx }) => {
    if (ctx.role !== "admin") {
      throw new Error("Forbidden");
    }
  },
  handler: async () => ({ ok: true })
});

afterAction

Use afterAction to sanitize or shape the final response.

app/actions/users.ts

import { defineAction } from "@zapaction/core";
import { z } from "zod";

export const getUser = defineAction({
  input: z.object({ id: z.string() }),
  output: z.object({ id: z.string(), email: z.string() }),
  afterAction: async ({ result }) => ({
    ...result,
    email: result.email.toLowerCase()
  }),
  handler: async ({ input }) => ({
    id: input.id,
    email: "[e-mail]"
  })
});

Order of execution

1. input.parse
2. resolveActionContext
3. beforeAction
4. handler
5. afterAction
6. output.parse
7. revalidateTags