@openclaw/voice-call 2026.5.7 → 2026.5.10-beta.1

package/README.md

@@ -106,6 +106,8 @@ Notes:
 - advanced webhook, streaming, and tunnel notes: `https://docs.openclaw.ai/plugins/voice-call`
 - `responseModel` is optional. When unset, voice responses use the runtime default model.
 - `sessionScope` defaults to `per-phone`, preserving caller memory across calls. Use `per-call` for reception, booking, IVR, and bridge flows where each carrier call should start fresh.
+- `realtime.consultThinkingLevel` is optional. When set, it overrides the thinking level used by the model behind realtime `openclaw_agent_consult` calls.
+- `realtime.consultFastMode` is optional. When set, it toggles fast mode for realtime `openclaw_agent_consult` calls.
 
 ## Stale call reaper
 

package/dist/{config-7w04YpHh.js → config-D1yqfwMq.js}

@@ -189,6 +189,11 @@ const RealtimeToolSchema = z.object({
 }).strict();
 const VoiceCallRealtimeProvidersConfigSchema = z.record(z.string(), z.record(z.string(), z.unknown())).default({});
 const VoiceCallRealtimeToolPolicySchema = z.enum(REALTIME_VOICE_AGENT_CONSULT_TOOL_POLICIES);
+const VoiceCallRealtimeConsultPolicySchema = z.enum([
+	"auto",
+	"substantive",
+	"always"
+]);
 const VoiceCallRealtimeFastContextSourceSchema = z.enum(["memory", "sessions"]);
 const VoiceCallRealtimeFastContextConfigSchema = z.object({
 	/** Enable bounded memory/session lookup before the full consult agent. */
@@ -208,6 +213,45 @@ const VoiceCallRealtimeFastContextConfigSchema = z.object({
 	sources: ["memory", "sessions"],
 	fallbackToConsult: false
 });
+const VoiceCallRealtimeAgentContextConfigSchema = z.object({
+	/** Inject a compact agent persona/context capsule into realtime voice instructions. */
+	enabled: z.boolean().default(false),
+	/** Maximum number of characters from the generated capsule to append. */
+	maxChars: z.number().int().positive().default(6e3),
+	/** Include configured agent identity fields. */
+	includeIdentity: z.boolean().default(true),
+	/** Include agents.defaults/list systemPromptOverride when configured. */
+	includeSystemPrompt: z.boolean().default(true),
+	/** Include selected workspace files such as SOUL.md and IDENTITY.md. */
+	includeWorkspaceFiles: z.boolean().default(true),
+	/** Workspace-relative files to include, bounded by maxChars. */
+	files: z.array(z.string().min(1)).default([
+		"SOUL.md",
+		"IDENTITY.md",
+		"USER.md"
+	])
+}).strict().default({
+	enabled: false,
+	maxChars: 6e3,
+	includeIdentity: true,
+	includeSystemPrompt: true,
+	includeWorkspaceFiles: true,
+	files: [
+		"SOUL.md",
+		"IDENTITY.md",
+		"USER.md"
+	]
+});
+const VoiceCallRealtimeConsultThinkingLevelSchema = z.enum([
+	"off",
+	"minimal",
+	"low",
+	"medium",
+	"high",
+	"xhigh",
+	"adaptive",
+	"max"
+]);
 const VoiceCallStreamingProvidersConfigSchema = z.record(z.string(), z.record(z.string(), z.unknown())).default({});
 const VoiceCallRealtimeConfigSchema = z.object({
 	/** Enable realtime voice-to-voice mode. */
@@ -220,16 +264,25 @@ const VoiceCallRealtimeConfigSchema = z.object({
 	instructions: z.string().default(DEFAULT_VOICE_CALL_REALTIME_INSTRUCTIONS),
 	/** Tool policy for the shared OpenClaw agent consult tool. */
 	toolPolicy: VoiceCallRealtimeToolPolicySchema.default("safe-read-only"),
+	/** Guidance for when the realtime model should call the OpenClaw agent consult tool. */
+	consultPolicy: VoiceCallRealtimeConsultPolicySchema.default("auto"),
+	/** Optional thinking level override for the regular agent behind realtime consults. */
+	consultThinkingLevel: VoiceCallRealtimeConsultThinkingLevelSchema.optional(),
+	/** Optional fast mode override for the regular agent behind realtime consults. */
+	consultFastMode: z.boolean().optional(),
 	/** Tool definitions exposed to the realtime provider. */
 	tools: z.array(RealtimeToolSchema).default([]),
 	/** Low-latency memory/session context for the consult tool. */
 	fastContext: VoiceCallRealtimeFastContextConfigSchema,
+	/** Bounded agent persona/context injection for the fast realtime voice path. */
+	agentContext: VoiceCallRealtimeAgentContextConfigSchema,
 	/** Provider-owned raw config blobs keyed by provider id. */
 	providers: VoiceCallRealtimeProvidersConfigSchema
 }).strict().default({
 	enabled: false,
 	instructions: DEFAULT_VOICE_CALL_REALTIME_INSTRUCTIONS,
 	toolPolicy: "safe-read-only",
+	consultPolicy: "auto",
 	tools: [],
 	fastContext: {
 		enabled: false,
@@ -238,6 +291,18 @@ const VoiceCallRealtimeConfigSchema = z.object({
 		sources: ["memory", "sessions"],
 		fallbackToConsult: false
 	},
+	agentContext: {
+		enabled: false,
+		maxChars: 6e3,
+		includeIdentity: true,
+		includeSystemPrompt: true,
+		includeWorkspaceFiles: true,
+		files: [
+			"SOUL.md",
+			"IDENTITY.md",
+			"USER.md"
+		]
+	},
 	providers: {}
 });
 const VoiceCallStreamingConfigSchema = z.object({
@@ -423,6 +488,11 @@ function normalizeVoiceCallConfig(config) {
 		...config.realtime?.fastContext,
 		sources: config.realtime?.fastContext?.sources ?? defaults.realtime.fastContext.sources
 	};
+	const realtimeAgentContext = {
+		...defaults.realtime.agentContext,
+		...config.realtime?.agentContext,
+		files: config.realtime?.agentContext?.files ?? defaults.realtime.agentContext.files
+	};
 	return {
 		...defaults,
 		...config,
@@ -459,7 +529,10 @@ function normalizeVoiceCallConfig(config) {
 			provider: realtimeProvider,
 			streamPath: config.realtime?.streamPath ?? defaultRealtimeStreamPathForServePath(serve.path ?? defaults.serve.path),
 			tools: config.realtime?.tools ?? defaults.realtime.tools,
+			consultThinkingLevel: VoiceCallRealtimeConsultThinkingLevelSchema.optional().parse(config.realtime?.consultThinkingLevel ?? defaults.realtime.consultThinkingLevel),
+			consultFastMode: config.realtime?.consultFastMode ?? defaults.realtime.consultFastMode,
 			fastContext: realtimeFastContext,
+			agentContext: realtimeAgentContext,
 			providers: realtimeProviders
 		},
 		tts: normalizeVoiceCallTtsConfig(defaults.tts, config.tts)

package/dist/{config-compat-B0me39_4.js → config-compat-DphZfbFl.js}

@@ -1,4 +1,4 @@
-import { t as VoiceCallConfigSchema } from "./config-7w04YpHh.js";
+import { t as VoiceCallConfigSchema } from "./config-D1yqfwMq.js";
 import { asOptionalRecord, readStringField } from "openclaw/plugin-sdk/text-runtime";
 //#region extensions/voice-call/src/config-compat.ts
 const VOICE_CALL_LEGACY_CONFIG_REMOVAL_VERSION = "2026.6.0";

package/dist/{guarded-json-api-Btx5EE4w.js → guarded-json-api-C2JLyCeV.js}

@@ -2,6 +2,7 @@ import { fetchWithSsrFGuard } from "./runtime-api.js";
 import "./api.js";
 import { t as getHeader } from "./http-headers-BrnxBasF.js";
 import { formatErrorMessage } from "openclaw/plugin-sdk/error-runtime";
+import { isLoopbackHost } from "openclaw/plugin-sdk/gateway-runtime";
 import { normalizeLowercaseStringOrEmpty } from "openclaw/plugin-sdk/text-runtime";
 import crypto from "node:crypto";
 import { safeEqualSecret } from "openclaw/plugin-sdk/security-runtime";
@@ -194,12 +195,6 @@ function buildTwilioVerificationUrl(ctx, publicUrl, urlOptions) {
 		return publicUrl;
 	}
 }
-function isLoopbackAddress(address) {
-	if (!address) return false;
-	if (address === "127.0.0.1" || address === "::1") return true;
-	if (address.startsWith("::ffff:127.")) return true;
-	return false;
-}
 function stripPortFromUrl(url) {
 	try {
 		const parsed = new URL(url);
@@ -336,7 +331,7 @@ function verifyTwilioWebhook(ctx, authToken, options) {
 		ok: false,
 		reason: "Missing X-Twilio-Signature header"
 	};
-	const isLoopback = isLoopbackAddress(options?.remoteIP ?? ctx.remoteAddress);
+	const isLoopback = isLoopbackHost(options?.remoteIP ?? ctx.remoteAddress ?? "");
 	const allowLoopbackForwarding = options?.allowNgrokFreeTierLoopbackBypass && isLoopback;
 	const verificationUrl = buildTwilioVerificationUrl(ctx, options?.publicUrl, {
 		allowedHosts: options?.allowedHosts,

package/dist/index.js

@@ -1,8 +1,8 @@
 import { definePluginEntry, sleep } from "./runtime-api.js";
 import "./api.js";
-import { i as resolveVoiceCallConfig, s as validateProviderConfig } from "./config-7w04YpHh.js";
-import { a as setupTailscaleExposureRoute, i as getTailscaleSelfInfo, n as resolveWebhookExposureStatus, r as cleanupTailscaleExposureRoute, s as resolveUserPath, t as createVoiceCallRuntime } from "./runtime-entry-DFzuGKLG.js";
-import { i as parseVoiceCallPluginConfig, r as normalizeVoiceCallLegacyConfigInput, t as formatVoiceCallLegacyConfigWarnings } from "./config-compat-B0me39_4.js";
+import { i as resolveVoiceCallConfig, s as validateProviderConfig } from "./config-D1yqfwMq.js";
+import { a as setupTailscaleExposureRoute, i as getTailscaleSelfInfo, n as resolveWebhookExposureStatus, r as cleanupTailscaleExposureRoute, s as resolveUserPath, t as createVoiceCallRuntime } from "./runtime-entry-CVTYOHsM.js";
+import { i as parseVoiceCallPluginConfig, r as normalizeVoiceCallLegacyConfigInput, t as formatVoiceCallLegacyConfigWarnings } from "./config-compat-DphZfbFl.js";
 import { formatErrorMessage } from "openclaw/plugin-sdk/error-runtime";
 import { ErrorCodes, callGatewayFromCli, errorShape } from "openclaw/plugin-sdk/gateway-runtime";
 import { normalizeOptionalLowercaseString, normalizeOptionalString } from "openclaw/plugin-sdk/text-runtime";
@@ -732,6 +732,11 @@ const voiceCallConfigSchema = {
 			help: "Controls the shared openclaw_agent_consult tool.",
 			advanced: true
 		},
+		"realtime.consultPolicy": {
+			label: "Realtime Consult Policy",
+			help: "Guides when the realtime voice model should call openclaw_agent_consult.",
+			advanced: true
+		},
 		"realtime.fastContext.enabled": {
 			label: "Enable Fast Realtime Context",
 			help: "Searches memory/session context before the full consult agent.",
@@ -753,6 +758,31 @@ const voiceCallConfigSchema = {
 			label: "Fallback To Full Consult",
 			advanced: true
 		},
+		"realtime.agentContext.enabled": {
+			label: "Enable Agent Voice Context",
+			help: "Injects a compact agent identity, system prompt, and workspace context capsule into realtime voice instructions.",
+			advanced: true
+		},
+		"realtime.agentContext.maxChars": {
+			label: "Agent Voice Context Limit",
+			advanced: true
+		},
+		"realtime.agentContext.includeIdentity": {
+			label: "Include Agent Identity",
+			advanced: true
+		},
+		"realtime.agentContext.includeSystemPrompt": {
+			label: "Include Agent System Prompt",
+			advanced: true
+		},
+		"realtime.agentContext.includeWorkspaceFiles": {
+			label: "Include Agent Workspace Files",
+			advanced: true
+		},
+		"realtime.agentContext.files": {
+			label: "Agent Voice Context Files",
+			advanced: true
+		},
 		"realtime.providers": {
 			label: "Realtime Provider Config",
 			advanced: true
@@ -804,6 +834,8 @@ const VoiceCallToolSchema = Type.Union([
 		to: Type.Optional(Type.String({ description: "Call target" })),
 		message: Type.String({ description: "Intro message" }),
 		mode: Type.Optional(Type.Union([Type.Literal("notify"), Type.Literal("conversation")])),
+		sessionKey: Type.Optional(Type.String({ description: "OpenClaw session key for the call" })),
+		requesterSessionKey: Type.Optional(Type.String({ description: "OpenClaw session key that initiated the call" })),
 		dtmfSequence: Type.Optional(Type.String({ description: "DTMF digits to play before connect" }))
 	}),
 	Type.Object({
@@ -834,6 +866,8 @@ const VoiceCallToolSchema = Type.Union([
 		to: Type.Optional(Type.String({ description: "Call target" })),
 		sid: Type.Optional(Type.String({ description: "Call SID" })),
 		message: Type.Optional(Type.String({ description: "Optional intro message" })),
+		sessionKey: Type.Optional(Type.String({ description: "OpenClaw session key for the call" })),
+		requesterSessionKey: Type.Optional(Type.String({ description: "OpenClaw session key that initiated the call" })),
 		dtmfSequence: Type.Optional(Type.String({ description: "DTMF digits to play before connect" }))
 	})
 ]);
@@ -937,10 +971,11 @@ var voice_call_default = definePluginEntry({
 			return { error: await describeHistoricalCall(rt, callId) ?? "Call not found" };
 		};
 		const initiateCallAndRespond = async (params) => {
-			const result = await params.rt.manager.initiateCall(params.to, void 0, {
+			const result = await params.rt.manager.initiateCall(params.to, params.sessionKey, {
 				message: params.message,
 				mode: params.mode,
-				dtmfSequence: params.dtmfSequence
+				dtmfSequence: params.dtmfSequence,
+				...params.requesterSessionKey ? { requesterSessionKey: params.requesterSessionKey } : {}
 			});
 			if (!result.success) {
 				respondError(params.respond, result.error || "initiate failed");
@@ -985,7 +1020,9 @@ var voice_call_default = definePluginEntry({
 					respond,
 					to,
 					message,
-					mode: params?.mode === "notify" || params?.mode === "conversation" ? params.mode : void 0
+					mode: params?.mode === "notify" || params?.mode === "conversation" ? params.mode : void 0,
+					sessionKey: normalizeOptionalString(params?.sessionKey),
+					requesterSessionKey: normalizeOptionalString(params?.requesterSessionKey)
 				});
 			} catch (err) {
 				sendError(respond, err);
@@ -1128,6 +1165,8 @@ var voice_call_default = definePluginEntry({
 				const to = normalizeOptionalString(params?.to) ?? "";
 				const message = normalizeOptionalString(params?.message) ?? "";
 				const dtmfSequence = normalizeOptionalString(params?.dtmfSequence);
+				const sessionKey = normalizeOptionalString(params?.sessionKey);
+				const requesterSessionKey = normalizeOptionalString(params?.requesterSessionKey);
 				if (!to) {
 					respondError(respond, "to required", ErrorCodes.INVALID_REQUEST);
 					return;
@@ -1139,7 +1178,9 @@ var voice_call_default = definePluginEntry({
 					to,
 					message: message || void 0,
 					mode,
-					dtmfSequence
+					dtmfSequence,
+					sessionKey,
+					...requesterSessionKey ? { requesterSessionKey } : {}
 				});
 			} catch (err) {
 				sendError(respond, err);
@@ -1233,9 +1274,10 @@ var voice_call_default = definePluginEntry({
 					}
 					const to = normalizeOptionalString(rawParams.to) ?? rt.config.toNumber;
 					if (!to) throw new Error("to required for call");
-					const result = await rt.manager.initiateCall(to, void 0, {
+					const result = await rt.manager.initiateCall(to, normalizeOptionalString(rawParams.sessionKey), {
 						dtmfSequence: normalizeOptionalString(rawParams.dtmfSequence),
-						message: normalizeOptionalString(rawParams.message)
+						message: normalizeOptionalString(rawParams.message),
+						...normalizeOptionalString(rawParams.requesterSessionKey) ? { requesterSessionKey: normalizeOptionalString(rawParams.requesterSessionKey) } : {}
 					});
 					if (!result.success) throw new Error(result.error || "initiate failed");
 					return json({

package/dist/{plivo-B-a7KFoT.js → plivo-3Lcw84OQ.js}

@@ -1,6 +1,6 @@
 import { t as escapeXml } from "./voice-mapping-BYDGdWGx.js";
 import { t as getHeader } from "./http-headers-BrnxBasF.js";
-import { n as reconstructWebhookUrl, r as verifyPlivoWebhook, t as guardedJsonApiRequest } from "./guarded-json-api-Btx5EE4w.js";
+import { n as reconstructWebhookUrl, r as verifyPlivoWebhook, t as guardedJsonApiRequest } from "./guarded-json-api-C2JLyCeV.js";
 import { normalizeLowercaseStringOrEmpty, normalizeOptionalString } from "openclaw/plugin-sdk/text-runtime";
 import crypto from "node:crypto";
 //#region extensions/voice-call/src/providers/plivo.ts