Filehigh importancesource

hooks.ts

types/hooks.ts

No strong subsystem tag
291
Lines
9138
Bytes
16
Exports
11
Imports
10
Keywords

What this is

This page documents one file from the repository and includes its full source so you can read it without leaving the docs site.

Beginner explanation

This file is one piece of the larger system. Its name, directory, imports, and exports show where it fits. Start by reading the exports and related files first.

How it is used

Start from the exports list and related files. Those are the easiest clues for where this file fits into the system.

Expert explanation

Architecturally, this file intersects with general runtime concerns. It contains 291 lines, 11 detected imports, and 16 detected exports.

Important relationships

Detected exports

  • isHookEvent
  • promptRequestSchema
  • PromptRequest
  • PromptResponse
  • syncHookResponseSchema
  • hookJSONOutputSchema
  • isSyncHookJSONOutput
  • isAsyncHookJSONOutput
  • HookCallbackContext
  • HookCallback
  • HookCallbackMatcher
  • HookProgress
  • HookBlockingError
  • PermissionRequestResult
  • HookResult
  • AggregatedHookResult

Keywords

optionalobjectliteralhookeventnamemessageunknowndescribeadditionalcontextjsonhook

Detected imports

  • zod/v4
  • ../utils/lazySchema.js
  • src/entrypoints/agentSdkTypes.js
  • src/entrypoints/agentSdkTypes.js
  • src/types/message.js
  • src/utils/permissions/PermissionResult.js
  • src/utils/permissions/PermissionRule.js
  • src/utils/permissions/PermissionUpdateSchema.js
  • ../state/AppState.js
  • ../utils/commitAttribution.js
  • type-fest

Source notes

This page embeds the full file contents. Small or leaf files are still indexed honestly instead of being over-explained.

Open parent directory

Full source

// biome-ignore-all assist/source/organizeImports: ANT-ONLY import markers must not be reordered
import { z } from 'zod/v4'
import { lazySchema } from '../utils/lazySchema.js'
import {
  type HookEvent,
  HOOK_EVENTS,
  type HookInput,
  type PermissionUpdate,
} from 'src/entrypoints/agentSdkTypes.js'
import type {
  HookJSONOutput,
  AsyncHookJSONOutput,
  SyncHookJSONOutput,
} from 'src/entrypoints/agentSdkTypes.js'
import type { Message } from 'src/types/message.js'
import type { PermissionResult } from 'src/utils/permissions/PermissionResult.js'
import { permissionBehaviorSchema } from 'src/utils/permissions/PermissionRule.js'
import { permissionUpdateSchema } from 'src/utils/permissions/PermissionUpdateSchema.js'
import type { AppState } from '../state/AppState.js'
import type { AttributionState } from '../utils/commitAttribution.js'

export function isHookEvent(value: string): value is HookEvent {
  return HOOK_EVENTS.includes(value as HookEvent)
}

// Prompt elicitation protocol types. The `prompt` key acts as discriminator
// (mirroring the {async:true} pattern), with the id as its value.
export const promptRequestSchema = lazySchema(() =>
  z.object({
    prompt: z.string(), // request id
    message: z.string(),
    options: z.array(
      z.object({
        key: z.string(),
        label: z.string(),
        description: z.string().optional(),
      }),
    ),
  }),
)

export type PromptRequest = z.infer<ReturnType<typeof promptRequestSchema>>

export type PromptResponse = {
  prompt_response: string // request id
  selected: string
}

// Sync hook response schema
export const syncHookResponseSchema = lazySchema(() =>
  z.object({
    continue: z
      .boolean()
      .describe('Whether Claude should continue after hook (default: true)')
      .optional(),
    suppressOutput: z
      .boolean()
      .describe('Hide stdout from transcript (default: false)')
      .optional(),
    stopReason: z
      .string()
      .describe('Message shown when continue is false')
      .optional(),
    decision: z.enum(['approve', 'block']).optional(),
    reason: z.string().describe('Explanation for the decision').optional(),
    systemMessage: z
      .string()
      .describe('Warning message shown to the user')
      .optional(),
    hookSpecificOutput: z
      .union([
        z.object({
          hookEventName: z.literal('PreToolUse'),
          permissionDecision: permissionBehaviorSchema().optional(),
          permissionDecisionReason: z.string().optional(),
          updatedInput: z.record(z.string(), z.unknown()).optional(),
          additionalContext: z.string().optional(),
        }),
        z.object({
          hookEventName: z.literal('UserPromptSubmit'),
          additionalContext: z.string().optional(),
        }),
        z.object({
          hookEventName: z.literal('SessionStart'),
          additionalContext: z.string().optional(),
          initialUserMessage: z.string().optional(),
          watchPaths: z
            .array(z.string())
            .describe('Absolute paths to watch for FileChanged hooks')
            .optional(),
        }),
        z.object({
          hookEventName: z.literal('Setup'),
          additionalContext: z.string().optional(),
        }),
        z.object({
          hookEventName: z.literal('SubagentStart'),
          additionalContext: z.string().optional(),
        }),
        z.object({
          hookEventName: z.literal('PostToolUse'),
          additionalContext: z.string().optional(),
          updatedMCPToolOutput: z
            .unknown()
            .describe('Updates the output for MCP tools')
            .optional(),
        }),
        z.object({
          hookEventName: z.literal('PostToolUseFailure'),
          additionalContext: z.string().optional(),
        }),
        z.object({
          hookEventName: z.literal('PermissionDenied'),
          retry: z.boolean().optional(),
        }),
        z.object({
          hookEventName: z.literal('Notification'),
          additionalContext: z.string().optional(),
        }),
        z.object({
          hookEventName: z.literal('PermissionRequest'),
          decision: z.union([
            z.object({
              behavior: z.literal('allow'),
              updatedInput: z.record(z.string(), z.unknown()).optional(),
              updatedPermissions: z.array(permissionUpdateSchema()).optional(),
            }),
            z.object({
              behavior: z.literal('deny'),
              message: z.string().optional(),
              interrupt: z.boolean().optional(),
            }),
          ]),
        }),
        z.object({
          hookEventName: z.literal('Elicitation'),
          action: z.enum(['accept', 'decline', 'cancel']).optional(),
          content: z.record(z.string(), z.unknown()).optional(),
        }),
        z.object({
          hookEventName: z.literal('ElicitationResult'),
          action: z.enum(['accept', 'decline', 'cancel']).optional(),
          content: z.record(z.string(), z.unknown()).optional(),
        }),
        z.object({
          hookEventName: z.literal('CwdChanged'),
          watchPaths: z
            .array(z.string())
            .describe('Absolute paths to watch for FileChanged hooks')
            .optional(),
        }),
        z.object({
          hookEventName: z.literal('FileChanged'),
          watchPaths: z
            .array(z.string())
            .describe('Absolute paths to watch for FileChanged hooks')
            .optional(),
        }),
        z.object({
          hookEventName: z.literal('WorktreeCreate'),
          worktreePath: z.string(),
        }),
      ])
      .optional(),
  }),
)

// Zod schema for hook JSON output validation
export const hookJSONOutputSchema = lazySchema(() => {
  // Async hook response schema
  const asyncHookResponseSchema = z.object({
    async: z.literal(true),
    asyncTimeout: z.number().optional(),
  })
  return z.union([asyncHookResponseSchema, syncHookResponseSchema()])
})

// Infer the TypeScript type from the schema
type SchemaHookJSONOutput = z.infer<ReturnType<typeof hookJSONOutputSchema>>

// Type guard function to check if response is sync
export function isSyncHookJSONOutput(
  json: HookJSONOutput,
): json is SyncHookJSONOutput {
  return !('async' in json && json.async === true)
}

// Type guard function to check if response is async
export function isAsyncHookJSONOutput(
  json: HookJSONOutput,
): json is AsyncHookJSONOutput {
  return 'async' in json && json.async === true
}

// Compile-time assertion that SDK and Zod types match
import type { IsEqual } from 'type-fest'
type Assert<T extends true> = T
type _assertSDKTypesMatch = Assert<
  IsEqual<SchemaHookJSONOutput, HookJSONOutput>
>

/** Context passed to callback hooks for state access */
export type HookCallbackContext = {
  getAppState: () => AppState
  updateAttributionState: (
    updater: (prev: AttributionState) => AttributionState,
  ) => void
}

/** Hook that is a callback. */
export type HookCallback = {
  type: 'callback'
  callback: (
    input: HookInput,
    toolUseID: string | null,
    abort: AbortSignal | undefined,
    /** Hook index for SessionStart hooks to compute CLAUDE_ENV_FILE path */
    hookIndex?: number,
    /** Optional context for accessing app state */
    context?: HookCallbackContext,
  ) => Promise<HookJSONOutput>
  /** Timeout in seconds for this hook */
  timeout?: number
  /** Internal hooks (e.g. session file access analytics) are excluded from tengu_run_hook metrics */
  internal?: boolean
}

export type HookCallbackMatcher = {
  matcher?: string
  hooks: HookCallback[]
  pluginName?: string
}

export type HookProgress = {
  type: 'hook_progress'
  hookEvent: HookEvent
  hookName: string
  command: string
  promptText?: string
  statusMessage?: string
}

export type HookBlockingError = {
  blockingError: string
  command: string
}

export type PermissionRequestResult =
  | {
      behavior: 'allow'
      updatedInput?: Record<string, unknown>
      updatedPermissions?: PermissionUpdate[]
    }
  | {
      behavior: 'deny'
      message?: string
      interrupt?: boolean
    }

export type HookResult = {
  message?: Message
  systemMessage?: Message
  blockingError?: HookBlockingError
  outcome: 'success' | 'blocking' | 'non_blocking_error' | 'cancelled'
  preventContinuation?: boolean
  stopReason?: string
  permissionBehavior?: 'ask' | 'deny' | 'allow' | 'passthrough'
  hookPermissionDecisionReason?: string
  additionalContext?: string
  initialUserMessage?: string
  updatedInput?: Record<string, unknown>
  updatedMCPToolOutput?: unknown
  permissionRequestResult?: PermissionRequestResult
  retry?: boolean
}

export type AggregatedHookResult = {
  message?: Message
  blockingErrors?: HookBlockingError[]
  preventContinuation?: boolean
  stopReason?: string
  hookPermissionDecisionReason?: string
  permissionBehavior?: PermissionResult['behavior']
  additionalContexts?: string[]
  initialUserMessage?: string
  updatedInput?: Record<string, unknown>
  updatedMCPToolOutput?: unknown
  permissionRequestResult?: PermissionRequestResult
  retry?: boolean
}