@pikku/core 0.12.33 → 0.12.35

package/CHANGELOG.md

@@ -1,3 +1,26 @@
+## 0.12.35
+
+### Patch Changes
+
+- 6bca38f: Extend `aiAgentRunner` with AI SDK-style media methods for transcription, speech, image generation, embeddings, and reranking.
+
+  Move `voiceInput` and `voiceOutput` into `@pikku/core/ai-agent`, backed by the injected `aiAgentRunner`.
+
+  Deprecate `@pikku/ai-voice` and strip its exports.
+
+## 0.12.34
+
+### Patch Changes
+
+- 2eaa9fd: fix(workflow): seed sessionService when session already present on wire
+
+  When a parent workflow propagates its session to a child workflow via
+  `wire.session`, `resolveSession` skipped `setInitial` because `!wire.session`
+  was false, so `sessionService.freezeInitial()` returned `undefined` and
+  immediately overwrote the propagated session. We now seed the sessionService
+  with the existing `wire.session` so `freezeInitial()` returns the correct
+  session for `pikkuFunc` steps inside child workflows.
+
 ## 0.12.33
 
 ### Patch Changes

package/dist/index.d.ts

@@ -34,7 +34,7 @@ export type { SchemaService } from './services/schema-service.js';
 export type { SessionService } from './services/user-session-service.js';
 export { NoopAuditService, createInvocationAudit, resolveAuditActorFromWire, resolveAuditConfig, } from './services/audit-service.js';
 export type { AuditActor, AuditConfig, AuditDurability, AuditEvent, AuditEventBatch, AuditLog, AuditLogWriteInput, AuditOutcome, AuditService, AuditSource, ResolvedAuditConfig, } from './services/audit-service.js';
-export type { AIAgentRunnerService } from './services/ai-agent-runner-service.js';
+export type { AIAgentRunnerService, AIEmbedManyParams, AIEmbedManyResult, AIEmbedParams, AIEmbedResult, AIGenerateImageParams, AIGenerateImagePrompt, AIGenerateImageResult, AIGenerateSpeechParams, AIGenerateSpeechResult, AIProviderOptions, AIRerankParams, AIRerankResult, AITranscriptionParams, AITranscriptionResult, } from './services/ai-agent-runner-service.js';
 export type { AIRunStateService } from './services/ai-run-state-service.js';
 export type { AIStorageService } from './services/ai-storage-service.js';
 export type { EmailsMeta, EmailTemplateMeta, EmailTemplateLocaleMeta, EmailTemplateAssets, MetaService, } from './services/meta-service.js';

package/dist/services/ai-agent-runner-service.d.ts

@@ -1,4 +1,5 @@
 import type { AIMessage, AIAgentStep, AIAgentToolDef, AIStreamChannel } from '../wirings/ai-agent/ai-agent.types.js';
+export type AIProviderOptions = Record<string, Record<string, unknown>>;
 export type AIAgentRunnerParams = {
     model: string;
     temperature?: number;
@@ -41,9 +42,151 @@ export type AIAgentStepResult = {
     finishReason: 'stop' | 'tool-calls' | 'length' | 'error' | 'unknown';
     reasoningContent?: string;
 };
+export type AITranscriptionParams = {
+    model: string;
+    audio: Uint8Array;
+    providerOptions?: AIProviderOptions;
+    maxRetries?: number;
+    abortSignal?: AbortSignal;
+    headers?: Record<string, string>;
+};
+export type AITranscriptionResult = {
+    text: string;
+    segments?: Array<{
+        text: string;
+        startSecond: number;
+        endSecond: number;
+    }>;
+    language?: string;
+    durationInSeconds?: number;
+    warnings?: unknown[];
+    providerMetadata?: Record<string, unknown>;
+    responses?: unknown[];
+};
+export type AIGenerateSpeechParams = {
+    model: string;
+    text: string;
+    voice?: string;
+    outputFormat?: string;
+    instructions?: string;
+    speed?: number;
+    language?: string;
+    providerOptions?: AIProviderOptions;
+    maxRetries?: number;
+    abortSignal?: AbortSignal;
+    headers?: Record<string, string>;
+};
+export type AIGenerateSpeechResult = {
+    audio: {
+        uint8Array: Uint8Array;
+        base64: string;
+        mediaType: string;
+        format: string;
+    };
+    warnings?: unknown[];
+    providerMetadata?: Record<string, unknown>;
+    responses?: unknown[];
+};
+export type AIGenerateImagePrompt = string | {
+    images: Array<Uint8Array | ArrayBuffer | string>;
+    text?: string;
+    mask?: Uint8Array | ArrayBuffer | string;
+};
+export type AIGenerateImageParams = {
+    model: string;
+    prompt: AIGenerateImagePrompt;
+    n?: number;
+    maxImagesPerCall?: number;
+    size?: `${number}x${number}`;
+    aspectRatio?: `${number}:${number}`;
+    seed?: number;
+    providerOptions?: AIProviderOptions;
+    maxRetries?: number;
+    abortSignal?: AbortSignal;
+    headers?: Record<string, string>;
+};
+export type AIGenerateImageResult = {
+    images: Array<{
+        uint8Array: Uint8Array;
+        base64: string;
+        mediaType: string;
+    }>;
+    warnings?: unknown[];
+    providerMetadata?: Record<string, unknown>;
+    responses?: unknown[];
+    usage?: {
+        inputTokens?: number;
+        outputTokens?: number;
+        totalTokens?: number;
+    };
+};
+export type AIEmbedParams = {
+    model: string;
+    value: string;
+    providerOptions?: AIProviderOptions;
+    maxRetries?: number;
+    abortSignal?: AbortSignal;
+    headers?: Record<string, string>;
+};
+export type AIEmbedResult = {
+    value: string;
+    embedding: number[];
+    usage?: {
+        tokens?: number;
+    };
+    warnings?: unknown[];
+    providerMetadata?: Record<string, unknown>;
+    response?: unknown;
+};
+export type AIEmbedManyParams = {
+    model: string;
+    values: string[];
+    providerOptions?: AIProviderOptions;
+    maxRetries?: number;
+    abortSignal?: AbortSignal;
+    headers?: Record<string, string>;
+    maxParallelCalls?: number;
+};
+export type AIEmbedManyResult = {
+    values: string[];
+    embeddings: number[][];
+    usage?: {
+        tokens?: number;
+    };
+    warnings?: unknown[];
+    providerMetadata?: Record<string, unknown>;
+    responses?: unknown[];
+};
+export type AIRerankParams<VALUE extends string | Record<string, unknown>> = {
+    model: string;
+    query: string;
+    documents: VALUE[];
+    topK?: number;
+    providerOptions?: AIProviderOptions;
+    maxRetries?: number;
+    abortSignal?: AbortSignal;
+    headers?: Record<string, string>;
+};
+export type AIRerankResult<VALUE extends string | Record<string, unknown>> = {
+    ranking: Array<{
+        index: number;
+        document: VALUE;
+        score: number;
+    }>;
+    rerankedDocuments: VALUE[];
+    originalDocuments: VALUE[];
+    providerMetadata?: Record<string, unknown>;
+    response?: unknown;
+};
 export interface AIAgentRunnerService {
     stream(params: AIAgentRunnerParams, channel: AIStreamChannel): Promise<AIAgentStepResult>;
     run(params: AIAgentRunnerParams): Promise<AIAgentStepResult>;
+    transcribe?(params: AITranscriptionParams): Promise<AITranscriptionResult>;
+    generateSpeech?(params: AIGenerateSpeechParams): Promise<AIGenerateSpeechResult>;
+    generateImage?(params: AIGenerateImageParams): Promise<AIGenerateImageResult>;
+    embed?(params: AIEmbedParams): Promise<AIEmbedResult>;
+    embedMany?(params: AIEmbedManyParams): Promise<AIEmbedManyResult>;
+    rerank?<VALUE extends string | Record<string, unknown>>(params: AIRerankParams<VALUE>): Promise<AIRerankResult<VALUE>>;
     /** Return a new runner that uses the given API key for every LLM call.
      *  Optional — runners that don't support per-key scoping leave this undefined. */
     withApiKey?(apiKey: string): AIAgentRunnerService;

package/dist/services/index.d.ts

@@ -30,7 +30,7 @@ export type { TriggerService } from './trigger-service.js';
 export type { GatewayService } from './gateway-service.js';
 export type { DeploymentService, DeploymentConfig, DeploymentInfo, DeploymentServiceConfig, } from './deployment-service.js';
 export type { AIStorageService } from './ai-storage-service.js';
-export type { AIAgentRunnerParams, AIAgentRunnerResult, AIAgentStepResult, AIAgentRunnerService, } from './ai-agent-runner-service.js';
+export type { AIAgentRunnerParams, AIAgentRunnerResult, AIAgentStepResult, AIEmbedManyParams, AIEmbedManyResult, AIEmbedParams, AIEmbedResult, AIGenerateImageParams, AIGenerateImagePrompt, AIGenerateImageResult, AIGenerateSpeechParams, AIGenerateSpeechResult, AIProviderOptions, AIRerankParams, AIRerankResult, AITranscriptionParams, AITranscriptionResult, AIAgentRunnerService, } from './ai-agent-runner-service.js';
 export type { CreateRunInput, AIRunStateService, } from './ai-run-state-service.js';
 export type { CredentialStatus, CredentialMeta, } from './typed-secret-service.js';
 export type { CredentialService } from './credential-service.js';

package/dist/wirings/ai-agent/index.d.ts

@@ -1,6 +1,8 @@
 export { agent, agentStream, agentResume, agentApprove, } from './ai-agent-helpers.js';
 export { runAIAgent, resumeAIAgentSync } from './ai-agent-runner.js';
 export { streamAIAgent, resumeAIAgent } from './ai-agent-stream.js';
+export { voiceInput } from './voice-input.js';
+export { voiceOutput } from './voice-output.js';
 export { type RunAIAgentParams, type StreamAIAgentOptions, ToolApprovalRequired, ToolCredentialRequired, } from './ai-agent-prepare.js';
 export { addAIAgent, approveAIAgent, getAIAgents, getAIAgentsMeta, } from './ai-agent-registry.js';
 export type { AIAgentInput, AIAgentInputAttachment, AIAgentMeta, AIAgentMemoryConfig, AIAgentStep, AIContentPart, AgentRunRow, AgentRunService, AgentRunState, AIMessage, AIStreamChannel, AIStreamEvent, AIThread, CoreAIAgent, PendingApproval, PikkuAIMiddlewareHooks, } from './ai-agent.types.js';

package/dist/wirings/ai-agent/index.js

@@ -1,5 +1,7 @@
 export { agent, agentStream, agentResume, agentApprove, } from './ai-agent-helpers.js';
 export { runAIAgent, resumeAIAgentSync } from './ai-agent-runner.js';
 export { streamAIAgent, resumeAIAgent } from './ai-agent-stream.js';
+export { voiceInput } from './voice-input.js';
+export { voiceOutput } from './voice-output.js';
 export { ToolApprovalRequired, ToolCredentialRequired, } from './ai-agent-prepare.js';
 export { addAIAgent, approveAIAgent, getAIAgents, getAIAgentsMeta, } from './ai-agent-registry.js';

package/dist/wirings/ai-agent/voice-input.d.ts

@@ -0,0 +1,9 @@
+export declare const voiceInput: (config?: {
+    language?: string;
+    model?: string;
+    allowedAudioHosts?: string[];
+}) => import("./ai-agent.types.js").PikkuAIMiddlewareHooks<Record<string, unknown>, import("../../types/core.types.js").CoreSingletonServices<{
+    logLevel?: import("../../services/logger.js").LogLevel;
+    secrets?: {};
+    workflow?: import("../workflow/workflow.types.js").WorkflowServiceConfig;
+}>>;

package/dist/wirings/ai-agent/voice-input.js

@@ -0,0 +1,113 @@
+import { pikkuAIMiddleware } from '../../types/core.types.js';
+function base64ToUint8Array(base64) {
+    const binary = atob(base64);
+    const bytes = new Uint8Array(binary.length);
+    for (let i = 0; i < binary.length; i++) {
+        bytes[i] = binary.charCodeAt(i);
+    }
+    return bytes;
+}
+const MAX_AUDIO_SIZE = 50 * 1024 * 1024;
+// Portable SSRF guard: @pikku/core runs in edge runtimes (CF Workers) with no
+// Node `dns`, so we cannot resolve hostnames to check for private targets.
+// Reject the obvious internal literals; callers wanting stricter control pass
+// an explicit `allowedAudioHosts` allowlist. (Does not defend against a public
+// hostname that resolves to a private IP / DNS rebinding — out of reach here.)
+function isPrivateHost(hostname) {
+    const host = hostname.replace(/^\[|\]$/g, '').toLowerCase();
+    if (host === 'localhost' || host === '0.0.0.0' || host === '::1')
+        return true;
+    if (host.startsWith('fe80:') || host.startsWith('fc') || host.startsWith('fd'))
+        return true;
+    const v4 = host.match(/^(\d{1,3})\.(\d{1,3})\.(\d{1,3})\.\d{1,3}$/);
+    if (v4) {
+        const [a, b] = [Number(v4[1]), Number(v4[2])];
+        if (a === 127 || a === 10 || a === 0)
+            return true;
+        if (a === 169 && b === 254)
+            return true; // link-local incl. cloud metadata
+        if (a === 172 && b >= 16 && b <= 31)
+            return true;
+        if (a === 192 && b === 168)
+            return true;
+    }
+    return false;
+}
+async function fetchAsUint8Array(url, allowedAudioHosts) {
+    const parsed = new URL(url);
+    if (parsed.protocol !== 'https:' && parsed.protocol !== 'http:') {
+        throw new Error('Only HTTP(S) URLs are supported for audio');
+    }
+    if (allowedAudioHosts) {
+        if (!allowedAudioHosts.includes(parsed.hostname)) {
+            throw new Error(`Audio URL host is not allowed: ${parsed.hostname}`);
+        }
+    }
+    else if (isPrivateHost(parsed.hostname)) {
+        throw new Error(`Refusing to fetch audio from a private/internal host: ${parsed.hostname}`);
+    }
+    const response = await fetch(url);
+    const contentLength = response.headers.get('content-length');
+    if (contentLength && parseInt(contentLength, 10) > MAX_AUDIO_SIZE) {
+        throw new Error('Audio file exceeds maximum size');
+    }
+    const buffer = await response.arrayBuffer();
+    if (buffer.byteLength > MAX_AUDIO_SIZE) {
+        throw new Error('Audio file exceeds maximum size');
+    }
+    return new Uint8Array(buffer);
+}
+export const voiceInput = (config) => pikkuAIMiddleware({
+    modifyInput: async (services, { messages, instructions }) => {
+        const transcribeAudio = services.aiAgentRunner?.transcribe;
+        if (!transcribeAudio)
+            return { messages, instructions };
+        const last = messages[messages.length - 1];
+        if (!last || last.role !== 'user' || typeof last.content === 'string') {
+            return { messages, instructions };
+        }
+        const parts = last.content;
+        if (!parts)
+            return { messages, instructions };
+        const hasAudio = parts.some((p) => p.type === 'file' && !!p.mediaType?.startsWith('audio/'));
+        if (!hasAudio)
+            return { messages, instructions };
+        // Process parts in order, replacing each audio part with its transcription
+        // in place. Sequential, so no unbounded parallel downloads/transcriptions
+        // and original content ordering is preserved.
+        const updatedContent = [];
+        for (const p of parts) {
+            if (!(p.type === 'file' && p.mediaType?.startsWith('audio/'))) {
+                updatedContent.push(p);
+                continue;
+            }
+            if (!config?.model) {
+                throw new Error('voiceInput requires a transcription model (e.g. openai/whisper-1)');
+            }
+            const audioData = p.data
+                ? base64ToUint8Array(p.data)
+                : await fetchAsUint8Array(p.url, config.allowedAudioHosts);
+            const result = await transcribeAudio({
+                model: config.model,
+                audio: audioData,
+                ...(config.language
+                    ? {
+                        providerOptions: {
+                            openai: {
+                                language: config.language,
+                            },
+                        },
+                    }
+                    : {}),
+            });
+            updatedContent.push({ type: 'text', text: result.text });
+        }
+        return {
+            messages: [
+                ...messages.slice(0, -1),
+                { ...last, content: updatedContent },
+            ],
+            instructions,
+        };
+    },
+});

package/dist/wirings/ai-agent/voice-output.d.ts

@@ -0,0 +1,16 @@
+type VoiceOutputState = {
+    textBuffer?: string;
+};
+export declare const voiceOutput: (config?: {
+    model?: string;
+    format?: string;
+    voice?: string;
+    instructions?: string;
+    speed?: number;
+    language?: string;
+}) => import("./ai-agent.types.js").PikkuAIMiddlewareHooks<VoiceOutputState, import("../../types/core.types.js").CoreSingletonServices<{
+    logLevel?: import("../../services/logger.js").LogLevel;
+    secrets?: {};
+    workflow?: import("../workflow/workflow.types.js").WorkflowServiceConfig;
+}>>;
+export {};

package/dist/wirings/ai-agent/voice-output.js

@@ -0,0 +1,94 @@
+import { pikkuAIMiddleware } from '../../types/core.types.js';
+function bufferToBase64(data) {
+    let binary = '';
+    for (let i = 0; i < data.length; i++) {
+        binary += String.fromCharCode(data[i]);
+    }
+    return btoa(binary);
+}
+const SENTENCE_BOUNDARY = /[.!?]\s*$/;
+function isSentenceBoundary(text) {
+    return SENTENCE_BOUNDARY.test(text);
+}
+async function synthesizeAudio(aiAgentRunner, input) {
+    const result = await aiAgentRunner.generateSpeech?.({
+        model: input.model,
+        text: input.text,
+        voice: input.voice,
+        outputFormat: input.format,
+        instructions: input.instructions,
+        speed: input.speed,
+        language: input.language,
+    });
+    if (!result) {
+        throw new Error('voiceOutput requires an aiAgentRunner with generateSpeech support');
+    }
+    // Label chunks with the format the provider actually returned (config.format
+    // is only a request), falling back to the requested format then pcm16.
+    return {
+        bytes: result.audio.uint8Array,
+        format: result.audio.format || input.format || 'pcm16',
+    };
+}
+export const voiceOutput = (config) => pikkuAIMiddleware({
+    modifyOutputStream: async (services, { event, state }) => {
+        const aiAgentRunner = services.aiAgentRunner;
+        if (!aiAgentRunner?.generateSpeech)
+            return event;
+        if (event.type === 'done') {
+            const remaining = state.textBuffer ?? '';
+            if (remaining) {
+                state.textBuffer = '';
+                if (!config?.model) {
+                    throw new Error('voiceOutput requires a speech model (e.g. openai/tts-1)');
+                }
+                const audio = await synthesizeAudio(aiAgentRunner, {
+                    model: config.model,
+                    text: remaining,
+                    voice: config?.voice,
+                    format: config?.format,
+                    instructions: config?.instructions,
+                    speed: config?.speed,
+                    language: config?.language,
+                });
+                return [
+                    {
+                        type: 'audio-delta',
+                        data: bufferToBase64(audio.bytes),
+                        format: audio.format,
+                    },
+                    { type: 'audio-done' },
+                    event,
+                ];
+            }
+            return [{ type: 'audio-done' }, event];
+        }
+        if (event.type !== 'text-delta')
+            return event;
+        state.textBuffer = `${state.textBuffer ?? ''}${event.text}`;
+        if (!isSentenceBoundary(state.textBuffer))
+            return event;
+        const text = state.textBuffer;
+        state.textBuffer = '';
+        if (!config?.model) {
+            throw new Error('voiceOutput requires a speech model (e.g. openai/tts-1)');
+        }
+        const audio = await synthesizeAudio(aiAgentRunner, {
+            model: config.model,
+            text,
+            voice: config?.voice,
+            format: config?.format,
+            instructions: config?.instructions,
+            speed: config?.speed,
+            language: config?.language,
+        });
+        return [
+            event,
+            {
+                type: 'audio-delta',
+                data: bufferToBase64(audio.bytes),
+                format: audio.format,
+            },
+        ];
+    },
+});

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@pikku/core",
-  "version": "0.12.33",
+  "version": "0.12.35",
   "author": "yasser.fadl@gmail.com",
   "license": "MIT",
   "module": "dist/index.js",

package/src/index.ts

@@ -142,7 +142,23 @@ export type {
   AuditSource,
   ResolvedAuditConfig,
 } from './services/audit-service.js'
-export type { AIAgentRunnerService } from './services/ai-agent-runner-service.js'
+export type {
+  AIAgentRunnerService,
+  AIEmbedManyParams,
+  AIEmbedManyResult,
+  AIEmbedParams,
+  AIEmbedResult,
+  AIGenerateImageParams,
+  AIGenerateImagePrompt,
+  AIGenerateImageResult,
+  AIGenerateSpeechParams,
+  AIGenerateSpeechResult,
+  AIProviderOptions,
+  AIRerankParams,
+  AIRerankResult,
+  AITranscriptionParams,
+  AITranscriptionResult,
+} from './services/ai-agent-runner-service.js'
 export type { AIRunStateService } from './services/ai-run-state-service.js'
 export type { AIStorageService } from './services/ai-storage-service.js'
 export type {

package/src/services/ai-agent-runner-service.ts

@@ -5,6 +5,8 @@ import type {
   AIStreamChannel,
 } from '../wirings/ai-agent/ai-agent.types.js'
 
+export type AIProviderOptions = Record<string, Record<string, unknown>>
+
 export type AIAgentRunnerParams = {
   model: string
   temperature?: number
@@ -36,12 +38,169 @@ export type AIAgentStepResult = {
   reasoningContent?: string
 }
 
+export type AITranscriptionParams = {
+  model: string
+  audio: Uint8Array
+  providerOptions?: AIProviderOptions
+  maxRetries?: number
+  abortSignal?: AbortSignal
+  headers?: Record<string, string>
+}
+
+export type AITranscriptionResult = {
+  text: string
+  segments?: Array<{
+    text: string
+    startSecond: number
+    endSecond: number
+  }>
+  language?: string
+  durationInSeconds?: number
+  warnings?: unknown[]
+  providerMetadata?: Record<string, unknown>
+  responses?: unknown[]
+}
+
+export type AIGenerateSpeechParams = {
+  model: string
+  text: string
+  voice?: string
+  outputFormat?: string
+  instructions?: string
+  speed?: number
+  language?: string
+  providerOptions?: AIProviderOptions
+  maxRetries?: number
+  abortSignal?: AbortSignal
+  headers?: Record<string, string>
+}
+
+export type AIGenerateSpeechResult = {
+  audio: {
+    uint8Array: Uint8Array
+    base64: string
+    mediaType: string
+    format: string
+  }
+  warnings?: unknown[]
+  providerMetadata?: Record<string, unknown>
+  responses?: unknown[]
+}
+
+export type AIGenerateImagePrompt =
+  | string
+  | {
+      images: Array<Uint8Array | ArrayBuffer | string>
+      text?: string
+      mask?: Uint8Array | ArrayBuffer | string
+    }
+
+export type AIGenerateImageParams = {
+  model: string
+  prompt: AIGenerateImagePrompt
+  n?: number
+  maxImagesPerCall?: number
+  size?: `${number}x${number}`
+  aspectRatio?: `${number}:${number}`
+  seed?: number
+  providerOptions?: AIProviderOptions
+  maxRetries?: number
+  abortSignal?: AbortSignal
+  headers?: Record<string, string>
+}
+
+export type AIGenerateImageResult = {
+  images: Array<{
+    uint8Array: Uint8Array
+    base64: string
+    mediaType: string
+  }>
+  warnings?: unknown[]
+  providerMetadata?: Record<string, unknown>
+  responses?: unknown[]
+  usage?: {
+    inputTokens?: number
+    outputTokens?: number
+    totalTokens?: number
+  }
+}
+
+export type AIEmbedParams = {
+  model: string
+  value: string
+  providerOptions?: AIProviderOptions
+  maxRetries?: number
+  abortSignal?: AbortSignal
+  headers?: Record<string, string>
+}
+
+export type AIEmbedResult = {
+  value: string
+  embedding: number[]
+  usage?: { tokens?: number }
+  warnings?: unknown[]
+  providerMetadata?: Record<string, unknown>
+  response?: unknown
+}
+
+export type AIEmbedManyParams = {
+  model: string
+  values: string[]
+  providerOptions?: AIProviderOptions
+  maxRetries?: number
+  abortSignal?: AbortSignal
+  headers?: Record<string, string>
+  maxParallelCalls?: number
+}
+
+export type AIEmbedManyResult = {
+  values: string[]
+  embeddings: number[][]
+  usage?: { tokens?: number }
+  warnings?: unknown[]
+  providerMetadata?: Record<string, unknown>
+  responses?: unknown[]
+}
+
+export type AIRerankParams<VALUE extends string | Record<string, unknown>> = {
+  model: string
+  query: string
+  documents: VALUE[]
+  topK?: number
+  providerOptions?: AIProviderOptions
+  maxRetries?: number
+  abortSignal?: AbortSignal
+  headers?: Record<string, string>
+}
+
+export type AIRerankResult<VALUE extends string | Record<string, unknown>> = {
+  ranking: Array<{
+    index: number
+    document: VALUE
+    score: number
+  }>
+  rerankedDocuments: VALUE[]
+  originalDocuments: VALUE[]
+  providerMetadata?: Record<string, unknown>
+  response?: unknown
+}
+
 export interface AIAgentRunnerService {
   stream(
     params: AIAgentRunnerParams,
     channel: AIStreamChannel
   ): Promise<AIAgentStepResult>
   run(params: AIAgentRunnerParams): Promise<AIAgentStepResult>
+  transcribe?(params: AITranscriptionParams): Promise<AITranscriptionResult>
+  generateSpeech?(
+    params: AIGenerateSpeechParams
+  ): Promise<AIGenerateSpeechResult>
+  generateImage?(params: AIGenerateImageParams): Promise<AIGenerateImageResult>
+  embed?(params: AIEmbedParams): Promise<AIEmbedResult>
+  embedMany?(params: AIEmbedManyParams): Promise<AIEmbedManyResult>
+  rerank?<VALUE extends string | Record<string, unknown>>(
+    params: AIRerankParams<VALUE>
+  ): Promise<AIRerankResult<VALUE>>
   /** Return a new runner that uses the given API key for every LLM call.
    *  Optional — runners that don't support per-key scoping leave this undefined. */
   withApiKey?(apiKey: string): AIAgentRunnerService