Filehigh importancesource

permissions.ts

types/permissions.ts

442
Lines
13145
Bytes
31
Exports
2
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 shell-safety, permissions. It contains 442 lines, 2 detected imports, and 31 detected exports.

Important relationships

Detected exports

  • EXTERNAL_PERMISSION_MODES
  • ExternalPermissionMode
  • InternalPermissionMode
  • PermissionMode
  • INTERNAL_PERMISSION_MODES
  • PERMISSION_MODES
  • PermissionBehavior
  • PermissionRuleSource
  • PermissionRuleValue
  • PermissionRule
  • PermissionUpdateDestination
  • PermissionUpdate
  • WorkingDirectorySource
  • AdditionalWorkingDirectory
  • PermissionCommandMetadata
  • PermissionMetadata
  • PermissionAllowDecision
  • PendingClassifierCheck
  • PermissionAskDecision
  • PermissionDenyDecision
  • PermissionDecision
  • PermissionResult
  • PermissionDecisionReason
  • ClassifierResult
  • ClassifierBehavior
  • ClassifierUsage
  • YoloClassifierResult
  • RiskLevel
  • PermissionExplanation
  • ToolPermissionRulesBySource
  • ToolPermissionContext

Keywords

permissionclassifierstagereadonlyunknownwheninputreasonallowbehavior

Detected imports

  • bun:bundle
  • @anthropic-ai/sdk/resources/messages.mjs

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

/**
 * Pure permission type definitions extracted to break import cycles.
 *
 * This file contains only type definitions and constants with no runtime dependencies.
 * Implementation files remain in src/utils/permissions/ but can now import from here
 * to avoid circular dependencies.
 */

import { feature } from 'bun:bundle'
import type { ContentBlockParam } from '@anthropic-ai/sdk/resources/messages.mjs'

// ============================================================================
// Permission Modes
// ============================================================================

export const EXTERNAL_PERMISSION_MODES = [
  'acceptEdits',
  'bypassPermissions',
  'default',
  'dontAsk',
  'plan',
] as const

export type ExternalPermissionMode = (typeof EXTERNAL_PERMISSION_MODES)[number]

// Exhaustive mode union for typechecking. The user-addressable runtime set
// is INTERNAL_PERMISSION_MODES below.
export type InternalPermissionMode = ExternalPermissionMode | 'auto' | 'bubble'
export type PermissionMode = InternalPermissionMode

// Runtime validation set: modes that are user-addressable (settings.json
// defaultMode, --permission-mode CLI flag, conversation recovery).
export const INTERNAL_PERMISSION_MODES = [
  ...EXTERNAL_PERMISSION_MODES,
  ...(feature('TRANSCRIPT_CLASSIFIER') ? (['auto'] as const) : ([] as const)),
] as const satisfies readonly PermissionMode[]

export const PERMISSION_MODES = INTERNAL_PERMISSION_MODES

// ============================================================================
// Permission Behaviors
// ============================================================================

export type PermissionBehavior = 'allow' | 'deny' | 'ask'

// ============================================================================
// Permission Rules
// ============================================================================

/**
 * Where a permission rule originated from.
 * Includes all SettingSource values plus additional rule-specific sources.
 */
export type PermissionRuleSource =
  | 'userSettings'
  | 'projectSettings'
  | 'localSettings'
  | 'flagSettings'
  | 'policySettings'
  | 'cliArg'
  | 'command'
  | 'session'

/**
 * The value of a permission rule - specifies which tool and optional content
 */
export type PermissionRuleValue = {
  toolName: string
  ruleContent?: string
}

/**
 * A permission rule with its source and behavior
 */
export type PermissionRule = {
  source: PermissionRuleSource
  ruleBehavior: PermissionBehavior
  ruleValue: PermissionRuleValue
}

// ============================================================================
// Permission Updates
// ============================================================================

/**
 * Where a permission update should be persisted
 */
export type PermissionUpdateDestination =
  | 'userSettings'
  | 'projectSettings'
  | 'localSettings'
  | 'session'
  | 'cliArg'

/**
 * Update operations for permission configuration
 */
export type PermissionUpdate =
  | {
      type: 'addRules'
      destination: PermissionUpdateDestination
      rules: PermissionRuleValue[]
      behavior: PermissionBehavior
    }
  | {
      type: 'replaceRules'
      destination: PermissionUpdateDestination
      rules: PermissionRuleValue[]
      behavior: PermissionBehavior
    }
  | {
      type: 'removeRules'
      destination: PermissionUpdateDestination
      rules: PermissionRuleValue[]
      behavior: PermissionBehavior
    }
  | {
      type: 'setMode'
      destination: PermissionUpdateDestination
      mode: ExternalPermissionMode
    }
  | {
      type: 'addDirectories'
      destination: PermissionUpdateDestination
      directories: string[]
    }
  | {
      type: 'removeDirectories'
      destination: PermissionUpdateDestination
      directories: string[]
    }

/**
 * Source of an additional working directory permission.
 * Note: This is currently the same as PermissionRuleSource but kept as a
 * separate type for semantic clarity and potential future divergence.
 */
export type WorkingDirectorySource = PermissionRuleSource

/**
 * An additional directory included in permission scope
 */
export type AdditionalWorkingDirectory = {
  path: string
  source: WorkingDirectorySource
}

// ============================================================================
// Permission Decisions & Results
// ============================================================================

/**
 * Minimal command shape for permission metadata.
 * This is intentionally a subset of the full Command type to avoid import cycles.
 * Only includes properties needed by permission-related components.
 */
export type PermissionCommandMetadata = {
  name: string
  description?: string
  // Allow additional properties for forward compatibility
  [key: string]: unknown
}

/**
 * Metadata attached to permission decisions
 */
export type PermissionMetadata =
  | { command: PermissionCommandMetadata }
  | undefined

/**
 * Result when permission is granted
 */
export type PermissionAllowDecision<
  Input extends { [key: string]: unknown } = { [key: string]: unknown },
> = {
  behavior: 'allow'
  updatedInput?: Input
  userModified?: boolean
  decisionReason?: PermissionDecisionReason
  toolUseID?: string
  acceptFeedback?: string
  contentBlocks?: ContentBlockParam[]
}

/**
 * Metadata for a pending classifier check that will run asynchronously.
 * Used to enable non-blocking allow classifier evaluation.
 */
export type PendingClassifierCheck = {
  command: string
  cwd: string
  descriptions: string[]
}

/**
 * Result when user should be prompted
 */
export type PermissionAskDecision<
  Input extends { [key: string]: unknown } = { [key: string]: unknown },
> = {
  behavior: 'ask'
  message: string
  updatedInput?: Input
  decisionReason?: PermissionDecisionReason
  suggestions?: PermissionUpdate[]
  blockedPath?: string
  metadata?: PermissionMetadata
  /**
   * If true, this ask decision was triggered by a bashCommandIsSafe_DEPRECATED security check
   * for patterns that splitCommand_DEPRECATED could misparse (e.g. line continuations, shell-quote
   * transformations). Used by bashToolHasPermission to block early before splitCommand_DEPRECATED
   * transforms the command. Not set for simple newline compound commands.
   */
  isBashSecurityCheckForMisparsing?: boolean
  /**
   * If set, an allow classifier check should be run asynchronously.
   * The classifier may auto-approve the permission before the user responds.
   */
  pendingClassifierCheck?: PendingClassifierCheck
  /**
   * Optional content blocks (e.g., images) to include alongside the rejection
   * message in the tool result. Used when users paste images as feedback.
   */
  contentBlocks?: ContentBlockParam[]
}

/**
 * Result when permission is denied
 */
export type PermissionDenyDecision = {
  behavior: 'deny'
  message: string
  decisionReason: PermissionDecisionReason
  toolUseID?: string
}

/**
 * A permission decision - allow, ask, or deny
 */
export type PermissionDecision<
  Input extends { [key: string]: unknown } = { [key: string]: unknown },
> =
  | PermissionAllowDecision<Input>
  | PermissionAskDecision<Input>
  | PermissionDenyDecision

/**
 * Permission result with additional passthrough option
 */
export type PermissionResult<
  Input extends { [key: string]: unknown } = { [key: string]: unknown },
> =
  | PermissionDecision<Input>
  | {
      behavior: 'passthrough'
      message: string
      decisionReason?: PermissionDecision<Input>['decisionReason']
      suggestions?: PermissionUpdate[]
      blockedPath?: string
      /**
       * If set, an allow classifier check should be run asynchronously.
       * The classifier may auto-approve the permission before the user responds.
       */
      pendingClassifierCheck?: PendingClassifierCheck
    }

/**
 * Explanation of why a permission decision was made
 */
export type PermissionDecisionReason =
  | {
      type: 'rule'
      rule: PermissionRule
    }
  | {
      type: 'mode'
      mode: PermissionMode
    }
  | {
      type: 'subcommandResults'
      reasons: Map<string, PermissionResult>
    }
  | {
      type: 'permissionPromptTool'
      permissionPromptToolName: string
      toolResult: unknown
    }
  | {
      type: 'hook'
      hookName: string
      hookSource?: string
      reason?: string
    }
  | {
      type: 'asyncAgent'
      reason: string
    }
  | {
      type: 'sandboxOverride'
      reason: 'excludedCommand' | 'dangerouslyDisableSandbox'
    }
  | {
      type: 'classifier'
      classifier: string
      reason: string
    }
  | {
      type: 'workingDir'
      reason: string
    }
  | {
      type: 'safetyCheck'
      reason: string
      // When true, auto mode lets the classifier evaluate this instead of
      // forcing a prompt. True for sensitive-file paths (.claude/, .git/,
      // shell configs) — the classifier can see context and decide. False
      // for Windows path bypass attempts and cross-machine bridge messages.
      classifierApprovable: boolean
    }
  | {
      type: 'other'
      reason: string
    }

// ============================================================================
// Bash Classifier Types
// ============================================================================

export type ClassifierResult = {
  matches: boolean
  matchedDescription?: string
  confidence: 'high' | 'medium' | 'low'
  reason: string
}

export type ClassifierBehavior = 'deny' | 'ask' | 'allow'

export type ClassifierUsage = {
  inputTokens: number
  outputTokens: number
  cacheReadInputTokens: number
  cacheCreationInputTokens: number
}

export type YoloClassifierResult = {
  thinking?: string
  shouldBlock: boolean
  reason: string
  unavailable?: boolean
  /**
   * API returned "prompt is too long" — the classifier transcript exceeded
   * the context window. Deterministic (same transcript → same error), so
   * callers should fall back to normal prompting rather than retry/fail-closed.
   */
  transcriptTooLong?: boolean
  /** The model used for this classifier call */
  model: string
  /** Token usage from the classifier API call (for overhead telemetry) */
  usage?: ClassifierUsage
  /** Duration of the classifier API call in ms */
  durationMs?: number
  /** Character lengths of the prompt components sent to the classifier */
  promptLengths?: {
    systemPrompt: number
    toolCalls: number
    userPrompts: number
  }
  /** Path where error prompts were dumped (only set when unavailable due to API error) */
  errorDumpPath?: string
  /** Which classifier stage produced the final decision (2-stage XML only) */
  stage?: 'fast' | 'thinking'
  /** Token usage from stage 1 (fast) when stage 2 was also run */
  stage1Usage?: ClassifierUsage
  /** Duration of stage 1 in ms when stage 2 was also run */
  stage1DurationMs?: number
  /**
   * API request_id (req_xxx) for stage 1. Enables joining to server-side
   * api_usage logs for cache-miss / routing attribution. Also used for the
   * legacy 1-stage (tool_use) classifier — the single request goes here.
   */
  stage1RequestId?: string
  /**
   * API message id (msg_xxx) for stage 1. Enables joining the
   * tengu_auto_mode_decision analytics event to the classifier's actual
   * prompt/completion in post-analysis.
   */
  stage1MsgId?: string
  /** Token usage from stage 2 (thinking) when stage 2 was run */
  stage2Usage?: ClassifierUsage
  /** Duration of stage 2 in ms when stage 2 was run */
  stage2DurationMs?: number
  /** API request_id for stage 2 (set whenever stage 2 ran) */
  stage2RequestId?: string
  /** API message id (msg_xxx) for stage 2 (set whenever stage 2 ran) */
  stage2MsgId?: string
}

// ============================================================================
// Permission Explainer Types
// ============================================================================

export type RiskLevel = 'LOW' | 'MEDIUM' | 'HIGH'

export type PermissionExplanation = {
  riskLevel: RiskLevel
  explanation: string
  reasoning: string
  risk: string
}

// ============================================================================
// Tool Permission Context
// ============================================================================

/**
 * Mapping of permission rules by their source
 */
export type ToolPermissionRulesBySource = {
  [T in PermissionRuleSource]?: string[]
}

/**
 * Context needed for permission checking in tools
 * Note: Uses a simplified DeepImmutable approximation for this types-only file
 */
export type ToolPermissionContext = {
  readonly mode: PermissionMode
  readonly additionalWorkingDirectories: ReadonlyMap<
    string,
    AdditionalWorkingDirectory
  >
  readonly alwaysAllowRules: ToolPermissionRulesBySource
  readonly alwaysDenyRules: ToolPermissionRulesBySource
  readonly alwaysAskRules: ToolPermissionRulesBySource
  readonly isBypassPermissionsModeAvailable: boolean
  readonly strippedDangerousRules?: ToolPermissionRulesBySource
  readonly shouldAvoidPermissionPrompts?: boolean
  readonly awaitAutomatedChecksBeforeDialog?: boolean
  readonly prePlanMode?: PermissionMode
}